- Current Issue
- Accept Archive
-
XIAO Wenbo, LI Ao, WU Huaming, LI Yongbo, XIAO Bangzhi
2024,20(8):449-453, DOI: https://doi.org/10.1007/s11801-024-3194-2
Abstract:
This paper investigates the impedance spectroscopy of monocrystalline silicon solar cells (MSSC) and dye-sensitized solar cells (DSSC) using solar cell dynamic equivalent resistor-capacitance (RC) and capacitor sum capacitor (CSC) circuits. Firstly, these circuits effectively represent the dynamic behavior of MSSC and DSSC. Secondly, the measurement method significantly impacts the accuracy of impedance measurements in the high-frequency region. Finally, the series resistance affects the distance between the left end of the impedance spectroscopy and the origin, while the parallel resistance influences the size of the impedance spectroscopy. In the CSC circuit, the relative magnitudes of dielectric relaxation capacitance and chemical capacitance affect the number and position of arc in the impedance spectroscopy. The value of the dielectric relaxation capacitor determines the number of impedance spectroscopy arcs. These conclusions provide guidelines for improving the accuracy of solar cell impedance measurements.
-
Shaher Dwik, Gurusamy Sasikala
2024,20(8):454-459, DOI: https://doi.org/10.1007/s11801-024-3162-x
Abstract:
This paper presents the simulation and implementation of a reconfigurable pixel that serves both data acquisition and energy harvesting purposes. The main topic focuses on switching between the two operating modes of the photodiode:photoconductive and photovoltaic modes. This proposed model can be used to design novel optical sensors with energy harvesting capability, such as position sensitive device (PSD) and complementary metal oxide semiconductor (CMOS) image sensors, which can extend the battery lifetime of the whole optical system. Thus, we can overcome power supply problems like wiring and changing batteries frequently, especially in hard-to-reach places like space (cube satellites) or even underwater wireless optical communication (UWOC). The proposed pixel architecture offers the advantage of a minimalistic design with only four transistors. Nevertheless, it does come with a drawback in the form of higher noise levels. The simulation was achieved using MATLAB, and the implementation was performed using the programmable system-on-chip (PSoC) microcontroller. The results showed that the functionality of the dual-function pixel is correct, and the scheduling of both energy harvesting and signal sensing functions was successfully achieved.
-
2024,20(8):460-471, DOI: https://doi.org/10.1007/s11801-024-3164-8
Abstract:
Mid-infrared wavelength switchable and dual-wavelength random laser output has many potential applications. A polarization-maintaining random fiber Bragg grating (PMRFBG) array based on the photonic localization effect of longitudinal invariant transverse disorder in fiber structure is proposed, which can be used as random feedback of dual-wavelength and wavelength switchable output of random fiber laser (RFL). The random fiber Bragg grating (RFBG) array was designed on the panda-type polarization-maintaining fiber (PMF), and the two center wavelengths were 2 151.60 nm and 2 152.22 nm, respectively. The RFBG array was designed on the bow tie-type PMF, and the two center wavelengths were obtained, which were 2 153.08 nm and 2 153.96 nm, respectively. The RFBG array with a center wavelength of 2 139.27 nm was designed on single-mode fiber (SMF). The length of individual fiber Bragg grating (FBG) and PMRFBG, the refractive index modulation depth, the number of cascaded gratings, and the distance between gratings have different effects on the full width at half maximum (FWHM) and reflectance of the RFBG and PMRFBG array, but not on the central wavelength, as obtained by simulation using the transmission matrix method. The designed PMRFBG array provides theoretical support for the design of the feedback mechanism of RFL.
-
SUN Jinlu, QIU Yue, WU Yuhang, ZHAO Dan, MIAO Changyun
2024,20(8):472-476, DOI: https://doi.org/10.1007/s11801-024-3222-2
Abstract:
An algorithm that can implement particle shape recognition and size measurement in hybrid particle field is proposed. Based on the defocused images obtained by the interferometric particle imaging (IPI) system, shape recognition of particles can be realized through ResNet50. Two-dimensional (2D) Fourier transform and 2D autocorrelation transform are used to obtain the size of spherical and non-spherical particles, respectively. The shape and size of the particle in hybrid particle field are determined. Numerical simulation and experiment results suggest that the method has good accuracy in measuring the particle size in hybrid particle field.
-
LIU Qinpeng, WANG Danyang, LI Xingrui, GAO Hong, YU Dakuang
2024,20(8):477-482, DOI: https://doi.org/10.1007/s11801-024-3249-4
Abstract:
A temperature and refractive index sensor based on fiber Bragg grating (FBG) end surface cascade open Fabry-Pérot (FP) cavity has been designed and demonstrated experimentally. The open FP cavity has been fabricated on the end face of an FBG by dislocation fusion in this work, the open FP cavity could be used for refractive index sensing, and the temperature is measured by the FBG. The working principle of the sensor and the method of improving the sensitivity are analyzed by theoretical simulation. The refractive index sensitivity of the sensor is 1 108.4 nm/RIU, while the maximum fluctuation of the sensor stability experiment detection is 0.005 nm. The results show that it has satisfactory characteristics. The sensor is a compact all-fiber structure, so it has potential applications in the field of temperature refractive index sensing, such as biomedical and capacitor electrolyte detection.
-
WANG Jun, WANG Xuefei, ZHOU Boxiong, GUO Dongyan
2024,20(8):483-489, DOI: https://doi.org/10.1007/s11801-024-3172-8
Abstract:
Feature extraction of point clouds is a fundamental component of three-dimensional (3D) vision tasks. While existing feature extraction networks primarily focus on enhancing the geometric perception abilities of networks and overlook the crucial role played by coordinates. For instance, though two airplane wings share the same shape, it demands distinct feature representations due to their differing positions. In this paper, we introduce a novel module called position aware module (PAM) to leverage the coordinate features of points for positional encoding, and integrating this encoding into the feature extraction network to provide essential positional context. Furthermore, we embed PAM into the PointNet++ framework, and design a novel feature extraction network, named PointNetV3. To validate the effectiveness of PointNetV3, we conducted comprehensive experiments including classification, object tracking and object detection on point cloud. The results of remarkable improvement in three tasks demonstrate the exceptional performance achieved by PointNetV3 in point cloud processing.
-
ZHOU Liming, LI Shixin, ZHU Zhiren, CHEN Fankai, LIU Chen, DONG Xiuhuan
2024,20(8):490-496, DOI: https://doi.org/10.1007/s11801-024-3218-y
Abstract:
The traditional transmission line detection has the problems of low efficiency. To improve the performance, this paper proposes an improved you only look once version 5 (YOLOv5) transmission line foreign object detection algorithm. First, efficient channel attention (ECA) module is introduced in the backbone network for focusing the target features and improving the feature extraction capability of the network. Secondly, bilinear interpolation upsampling is introduced in the neck network to improve the model detection accuracy. Finally, by integrating the efficient intersection over union (EIoU) loss function and Soft non-maximum suppression (Soft NMS) algorithm, the convergence speed of the model is accelerated while the detection effect of the model is enhanced. Relative to the original algorithm, the improved algorithm reduces the number of parameters by 16.4%, increases the mean average precision (mAP)@0.5 by 3.9%, mAP@0.5:0.95 by 6.3%, and increases the detection speed to 55.3 frames per second (FPS). The improved algorithm is able to improve the performance of the foreign object detection in transmission lines effectively.
-
WANG Zhouliang, XIANG Wanni, WANG Weiya, LI Hui
2024,20(8):497-504, DOI: https://doi.org/10.1007/s11801-024-3255-6
Abstract:
Leaking data through screen-shooting has become the main way of modern leaks. Digital watermarking technology can trace the leaker through the watermark information after the data is leaked. The current screen-shooting watermarking scheme can resist part of the distortion in the screen-shooting process, but it faces two problems. On the one hand, the watermark capacity is small. On the other hand, when the shot watermarked image is incomplete, high watermark extraction accuracy cannot be guaranteed. Based on the above problems, we propose a resistant to incomplete shooting watermarking (RiSw) scheme. Specifically, we design a set of codecs that can embed binary images as watermarks into carrier images and extract them, which not only ensures good visual effects of watermarked image, but also greatly increases watermark capacity. To resist incomplete shooting, we propose an incomplete shooting layer to simulate the situation of incomplete shooting in the screen-shooting process. Robustness to incomplete shooting can be achieved through end-to-end training. Extensive experiments show that the scheme proposed in this paper has superior performance. Even if the watermarked image lacks 50% pixels, it can still maintain a stable extraction accuracy.
-
GE Yisu, YE Wenjie, ZHANG Guodao, LIN Mengying
2024,20(8):505-511, DOI: https://doi.org/10.1007/s11801-024-4139-5
Abstract:
With the deepening of neural network research, object detection has been developed rapidly in recent years, and video object detection methods have gradually attracted the attention of scholars, especially frameworks including multiple object tracking and detection. Most current works prefer to build the paradigm for multiple object tracking and detection by multi-task learning. Different with others, a multi-level temporal feature fusion structure is proposed in this paper to improve the performance of framework by utilizing the constraint of video temporal consistency. For training the temporal network end-to-end, a feature exchange training strategy is put forward for training the temporal feature fusion structure efficiently. The proposed method is tested on several acknowledged benchmarks, and encouraging results are obtained compared with the famous joint detection and tracking framework. The ablation experiment answers the problem of a good position for temporal feature fusion.
Volume 20,2024 Issue 8
-
Enhanced photoelectrochemical performance of TiO2/Sb2S3 nanorod arrays by annealing in Ar ambience
Abstract:
In this work, the TiO2/Sb2S3 nanorod arrays (NRAs) were synthesized through a two-stage hydrothermal route for photoelectrochemical (PEC) water splitting. The effect of annealing treatment in Ar ambience on the PEC activity of TiO2/Sb2S3 composite sample was investigated by electrochemical impedance analysis, including Nyquist and Mott-Schottky (M-S) plots. It was demonstrated that vacuum annealing could crystallize Sb2S3 component and change its color from red to black, leading to an increment of photocurrent density from 1.9 A/m2 to 4.25 A/m2 at 0 VSCE. The enhanced PEC performance was mainly attributed to the improved visible light absorption. Moreover, annealing treatment facilitated retarding the electron-hole recombination occurred at the solid/liquid interfaces. Our work might provide a novel strategy for enhancing the PEC performance of a semiconductor electrode.
- 1
-
Marine Organism Classification Method based on Hier-archical Multi-Scale Attention Mechanism
XV Haotian, Cheng Yuanzhi, Zhao Dong, Xie Peidong
Abstract:
We propose a hierarchical multi-scale attention mechanism-based model in response to the low accuracy and inefficient manual classification of existing oceanic biological image classification methods. Firstly, the H-EMA module is designed for lightweight feature extraction, achieving outstanding performance at a relatively low cost. Secondly, an improved Ef-ficientNetV2 Block is used to integrate information from different scales better and enhance inter-layer message passing. Furthermore, introducing the CBAM module enhances the model's perception of critical features, optimizing its generali-zation ability. Lastly, Focal Loss is introduced to adjust the weights of complex samples to address the issue of imbal-anced categories in the dataset, further improving the model's performance. The model achieved 96.11% accuracy on the intertidal marine organism dataset of Nanji Islands and 84.78% accuracy on the CIFAR-100 dataset, demonstrating its strong generalization ability to meet the demands of oceanic biological image classification.
- 1
-
Low-light image enhancement based on Multi-illumination estimation and Multi-scale fusion
Zhang Xinai, Gao Jing, Nie Kaiming, Luo Tao
Abstract:
To improve image quality under low illumination conditions, a novel low-light image enhancement method is proposed in this paper based on multi-illumination estimation and multi-scale fusion (MIMS). Firstly, the illumination is processed by contrast-limited adaptive histogram equalization (CLAHE), adaptive complementary gamma function (ACG), and adap-tive detail preserving S-curve (ADPS), respectively, to obtain three components. Then, the fusion-relevant features, exposure, and color contrast are selected as the weight maps. Subsequently, these components and weight maps are fused through multi-scale to generate enhanced illumination. Finally, the enhanced images are obtained by multiplying the en-hanced illumination and reflectance. Compared with existing approaches, this proposed method achieves an average in-crease of 0.81% and 2.89% in the SSIM and PSNR, and a decrease of 6.17% and 32.61% in the NIQE and GMSD, re-spectively.
- 1
-
Inverse design of broadband and dispersion-flattened highly GeO2-doped optical fibers based on neural networks and particle swarm algorithm
Abstract:
Reverse design of highly GeO2-doped silica optical fibers with broadband and flat dispersion profiles is proposed using a neural network combined with a particle swarm optimization algorithm. Firstly, the neural network model designed to predict optical fiber dispersion is trained with an appropriate choice of hyperparameters, achieving a root mean square error (RMSE) of 1.15?10-6 on the test dataset, with a determination coefficient (R-squared) of 0.999. Secondly, the neural network is combined with the particle swarm optimization algorithm for the inverse design of dispersion-flattened optical fibers. To expand the search space and avoid particles getting trapped in local optimal solutions, the particle swarm optimization algorithm incorporates adaptive inertia weight updating and a simulated annealing algorithm. Finally, by using a proper fitness function, the designed fibers exhibit flat group velocity dispersion (GVD) profiles at 1400-2400 nm, where the GVD fluctuations and minimum absolute GVD values are below 18 and 7 ps/nm?km, respectively.
- 1
-
Research on a compact and high sensitivity gas pressure sensor based on fiber Fabry-P?rot and Bragg grating*
Qinpeng Liu, Meihua Xing, Yang Di, Liu Bo
Abstract:
A compact and highly sensitive gas pressure and temperature sensor based on Fabry-P?rot interferometer(FPI) and fiber Bragg grating (FBG) is proposed and demonstrated experimentally in this letter. The theoretical model for pressure and temperature sensing is established to underpin the design. Building on this foundation, a novel micro silicon cavity sensor structure sensitive to pressure is devised downstream of the FBG. The concept of separate measurement and the mechanisms enhancing pressure sensitivity are meticulously analyzed, leading to the fabrication of corresponding samples. The sensor offers the advantages of compact size, robust construction, easy fabrication, and high sensitivity, making it potentially valuable for micro-pressure application.
- 1
-
Charatererization of High-Performance AlGaN-Based Solar-Blind UV Photodetectors
Fuyuting, zhan jie, liu bing, zheng fu, sun zhao lan
Abstract:
This study begins with the fabrication and simulation of high-performance back-illuminated AlGaN-based solar-blind UV photodetectors.Based on the photodetectors, a low-noise, high-gain UV detection system circuit is designed and fabricated, enabling the detection, acquisition, and calibration of weak solar-blind UV signals. Experimental results demonstrate that under zero bias conditions, with a UV light power density of 3.45 μW/cm2 at 260 nm, the sample achieves a peak responsivity (R) of 0.085 A/W, an external quantum efficiency (EQE) of 40.7%, and a detectivity (D*) of 7.46?1012 cm?Hz1/2?W-1. The system exhibits a bandpass characteristic within the 240~280 nm wavelength range, coupled with a high signal-to-noise ratio of 39.74 dB.
- 1
-
Multi-level temporal feature fusion with feature exchange strategy for multiple object tracking
Ge Yisu, YeWenjie, Zhang Guodao, Lin Mengying
Abstract:
With the deepening of neural network research, object detection has been developed rapidly in recent years, and video object detection methods have gradually attracted the attention of scholars, especially frameworks including multiple object tracking and detection. Most current works prefer to build the paradigm for multiple object tracking and detection by multi-task learning. Different with others, a multi-level temporal feature fusion structure is pro-posed in this paper to improve the performance of framework by utilizing the constraint of video temporal con-sistency. For training the temporal network end-to-end, a feature exchange training strategy is put forward for training the temporal feature fusion structure efficiently. The proposed method is tested on several acknowledged benchmarks, and encouraging results are obtained compared with the famous joint detection and tracking frame-work. The ablation experiment answers the problem of a good position for temporal feature fusion.
- 1
-
PCA-Net: A Heart segmentation model based on the meta-learning method
YangMengZhu, Zhu Dong, Dong Hao, HuShunBo, WangYongFang
Abstract:
In order to effectively prevent and treat heart-based diseases, the study of precise segmentation of heart parts is particularly important. The heart is divided into four parts: the left and right ventricles and the left and right atria, and the left main trunk is more important, thus the left ventricular muscle (LV-MYO),which is located in the middle part of the heart, has become the object of many researches. Deep learning medical image segmentation methods become the main means of image analysis and processing at present, but the deep learning methods based on traditional convolutional neural network (CNN) are not suitable for segmenting organs with few labels and few samples like the heart, while the metalearning methods are able to solve the above problems and achieve better results in the direction of heart segmentation. Since the LV-MYO is wrapped in the left ventricular blood pool (LV-BP), this paper proposes a new model for heart segmentation: PCA-Net. Specifically, we redesign the coding structure of Q-Net and make improvements in threshold extraction. Experimental results confirm that PCA-Net effectively improves the accuracy of segmenting LV-MYO and LVBP sites on the CMR dataset, and is validated on another publicly available dataset, ABD, where the results outperform other SOTA methods.
- 1
-
Preparation of water-soluble fluorescent probe based on copper nanoparticles and its detection of tetracycline in meat
ZULPIYE Hasanjan, DOU Xiao-zong, Zhang hongyan
Abstract:
In this study, a water-soluble fluorescent probe based on copper nanoparticles (CuNPs) was synthesized using a hydrothermal method with BSA (BSA@CuNPs) as a protectant, and it was utilized to efficiently detect the concentration of tetracycline (TC). Experiments show that the emission wavelength of the BSA@CuNPs probe, excited at 328 nm, was measured to be 403 nm. Upon the addition of TC to the BSA@CuNPs probe, its fluorescence intensity at 403 nm is obviously diminished due to the keto-enol functions C=O group in TC connects with the hydroxyl group on the BSA@CuNPs probe to form a non-fluorescent complex, which can be successfully utilized for detecting TC concentration. Additionally, a good linear relationship was observed between relative fluorescence intensity (F0/F) of BSA@CuNPs probe and the TC concentration from 20 to 130 mM, and the limit of detection (LOD) was 60 nM. Furthermore, the BSA@CuNPs probe was applied to detect TC concentrations in real meat samples, with recoveries ranging from 95.4% to 102.1%, and the relative standard deviation was below 2.7%. Therefore, BSA@CuNPs probe exhibits high sensitivity and selectivity for TC, which will be used for the study of TC analysis in meat.
- 1
-
Performance Enhancement of Perovskite Solar Cells with Cs2PtI6-xBrx Absorber Layer based on Numerical Simulation
Duzuyan, yjyuan, Houguofu, Liwei, Bijinlian, Xingyupeng
Abstract:
Cs2TiBr6, characterized as a lead-free vacancy-ordered double perovskite material, demonstrates notable stability and photoelectric performance, rendering it suitable for application in perovskite solar cells. However, the experimental efficiency of this material is currently limited, achieving only 3.12%. This efficiency limitation primarily stems from the material’s wide bandgap, which results in insufficient light absorption. To address this issue, this study utilized Cs2PtI6-xBrx for band structure fine-tuning. By adjusting the bandgap of Cs2PtI6-xBrx through Br- doping, not only was the alignment of the energy bandgap in the absorption layer effectively adjusted, but the cutoff range of light absorption was also significantly expanded. Setting the Cs2PtI6-xBrx bandgap to 1.4 eV resulted in optimal device performance. Additionally, comparative analyses aimed at boosting solar cell efficiency were performed on both hole transport and electron transport layer materials. Subsequently, the structure of the perovskite solar cell was optimized using an FTO/TiO2/Cs2TiBr6/Cs2PtI6-xBrx/Cu2O/Au configuration. Furthermore, the influence of the absorber layer thickness and defect density on device performance was studied. Ultimately, this configuration enabled the Cs2TiBr6/Cs2PtI6-xBrx solar cell to achieve a peak conversion efficiency of 29.69%.
- 1
-
MDCFS: Few-shot Object Detection with Multilevel decoupled Classifiers
LI Chuanhui, ZHOU Mian, FENG Zhikun, SUI Zezhou, GAO Zan
Abstract:
Few-shot object detection poses unique challenges as it requires effectively learning novel classes with limited labeled data. Current approaches often suffer from biases towards base classes during fine-tuning, leading to suboptimal performance on detecting novel classes. Additionally, in complex scenes, the confusion between foreground and background objects further affects the accuracy and robustness of the model. To address these issues, we propose the Multilevel De-coupling Classification Few-Shot Algorithm (MDCFS). we decouple the standard classifier into a parallel foreground classifier and a background classifier in the Few-Shot Object Detection (FSOD) setting. This decoupling enables the independent separation of positive samples from noisy negative samples, alleviating the foreground-background confusion problem commonly encountered in few-shot detectors. For Generalized Few-Shot Object Detection (G-FSOD), where the few-shot dataset contains base classes, we further decouple the foreground classification head into a base class classification head and a novel class classification head. To ensure balance, we assign more weight to the novel class classification head, effectively addressing the bias towards base classes. Furthermore, we optimize the initial weights of the few-shot fine-tuning stage, significantly reducing training time and mitigating catastrophic forgetting in G-FSOD. Additionally, we incorporate metric learning into our model with minimal cost. Experimental results demonstrate the effectiveness of our approach. Compared to state-of-the-art few-shot object detection methods based on fine-tuning, MDCFS achieves performance improvements of up to 6.3% on the PASCAL VOC dataset and 1.5% on the COCO dataset.
- 1
-
Optimization of CZTSSe thin film solar cell performance using PbS back surface field layer
luowenjing, baizhijian, xueyuming, wangluoxin, dingyanhong, daihongli
Abstract:
CZTSSe is an Earth-abundant, environmentally friendly, and low-cost material, with a current record power conversion efficiency of 14.9%, This result is well below the prediction of the Shockley-Queisser limit. To improve the performance of solar cells, a layer of PbS (BSF) was added to the CZTSSe, which was simulated using the SCAPS-1D software. The optimal photovoltaic performance of AZO/ZnO/CdS/CZTSSe/PbS/Mo was studied by changing the parameters of the CZTSSe absorption layer and PbS layer. Studies have shown that the efficiency can be increased to 20.86% after using PbS.
- 1
-
Traffic safety helmet wear detection based on improved YOLOv5 network
Abstract:
Aiming at the problem that the current traffic safety helmet detection model can’t balance the accuracy of detection with the size of the model and the poor generalization of the model, a method based on improving YOLOv5 is proposed. By incorporating the lightweight GhostNet module into the YOLOv5 backbone network, we effectively reduce the model size. The addition of the Receptive Fields Block (RFB) module enhances feature extraction and improves the feature acquisition capability of the lightweight model. Subsequently, the high-performance light-weight convolution, GSConv, is integrated into the neck structure for further model size compression. Moreover, the baseline model's loss function is substituted with EIoU, accelerating network convergence and enhancing de-tection precision. Experimental results corroborate the effectiveness of this improved algorithm in real-world traffic scenarios.
- 1
-
Actively tunable electromagnetically induced transparency in hybrid Dirac-VO2 metamaterials
Abstract:
In this paper, we present a metamaterial structure of Dirac and vanadium dioxide and investigate its optical properties using the finite-difference time-domain (FDTD) technique. Using the phase transition feature of vanadium dioxide, the design can realize active tuning of the PIT effect at terahertz frequency, thereby converting from a single PIT to a double PIT. When VO2 is in the insulating state, the structure is symmetric to obtain a single-band PIT effect; When VO2 is in the metallic state, the structure turns asymmetric to realize a dual-band PIT effect. This design provides a reference direction for the design of actively tunable metamaterials. Additionally, it is discovered that the transparent window's resonant frequency and the Dirac material's Fermi level in this structure have a somewhat linear relationship. In addition, the structure achieves superior refractive index sensitivity in the terahertz band, surpassing 1 THz/RIU. Consequently, the concept exhibits encouraging potential for application in refractive index sensors and optical switches.
- 1
-
Crosstalk-free multichannel sensor based on a dual-Sagnac interferometer interrogated by microwave photonic filter
Zhang Ailing, Wang Bo, Yongfu Zhang, Pengxiang Chang, Minghe Wang, Rui Lin
Abstract:
A crosstalk-free multichannel sensor based on a dual-Sagnac interferometer interrogated by microwave photonic filter (MPF) is demonstrated theoretically and experimentally in this paper. The dual-Sagnac interferometer is taken as the sensing element as well as an optical spectrum slicer to implement a dual-passband MPF. Experimental results show that the proposed multichannel sensor is crosstalk-free, and the temperature sensitivity of the each channel is -0.9884 MHz/℃ and -1.9302 MHz/℃, respectively. The sensing interrogation system has good linearity and stability and it is suitable for application scenarios where multiple points of temperature are required to monitor simultaneously.
- 1
-
Terahertz dual-beam leaky-wave antenna based on multi-mode spoof surface plasmon polariton waveguide
Abstract:
Abstract: In this paper, a terahertz dual-beam leaky-wave antenna (LWA) based on a multi-mode spoof surface plasmon polariton (SSPP) waveguide is proposed. By dispersion engineering of the SSPP transmission line (SSPP TL), multiple modes are excited on the SSPP TL. A dual-beam LWA is constructed based on the radiation from the first space harmonics of the guided fundamental mode and the first higher-order mode. The simulation results show that the antenna radiates frequency-dependent dual beams in the 1.9-2.6 THz operating frequency band. The average gains for the two beams are 12.9 dBi and 8.2 dBi, respectively. The compact dual-beam LWA is promising in the application of wireless communication and radar systems.
- 1
-
Design, Fabrication and Characteristics of Optofluidic Variable Aperture based on Electromagnetic-driving
zhang anning, zhao rui, wei xian, liang zhongcheng, han zefeng
Abstract:
An electromagnetic-driving variable liquid aperture is designed and fabricated. It consists of a driving cavity, an optical cavity and a storage chamber, a polydimethylsiloxane（PDMS）elastic film, an annular magnet and a driving coils. The driving cavity is filled with dyeing liquid while the colorless transparent liquid fills the optical cavity and the storage chamber. When applying current, the annular magnet moves downward driven by the magnetic field leading to deformation of the PDMS film. Being squeezed by the moving magnet with the deforming film, the dyeing liquid flows from the driving cavity into the optical cavity, which contributes to the decrease of the clear aperture diameter in the optical cavity. Our proposed aperture performs continuous variable ability up to 2.775 mm in diameter of clear aperture when the current increases from 0 to 0.33A, and its relative transmittance ranges from 98.539% to 22.776% ~390 nm.
- 1
-
Representation Strategy for Unsupervised Domain Adaptation on Person Re-Identification
Hao Li, Tao Zhang, Shuang Li, Xuan Li, Xin Zhao
Abstract:
The task of unsupervised person re-identification (Re-ID) is to transfer the knowledge learned in the source domain with no labels to the target domain with no labels. Due to the significant differences in the background of different datasets, the trained model is challenging to extract person features accurately on unsupervised domain adaptive (UDA). Most UDA methods for person re-ID use single-image representation (SIR) during the feature extraction. These methods might ignore the difference among the cross-view images with the same identity. For this problem, the Joint Learning Image Representation Strategy for Unsupervised Domain Adaptation (JLIRS-UDA), which takes cross-image representation (CIR) into account for UDA, is proposed. The network architecture of JLIRS-UDA consists of two networks with branching networks. Each network consists of a shared network and two branching networks the SIR branch and CIR branch. The two branching networks aim to learn the SIR and CIR, respectively. To ensure the accuracy of the pseudo-label generation, the Segmenting Dynamic Clustering (SDC) method, which divides the training process into two phases, is proposed. Precisely, in the first phases, SDC adopts the single image features in the clustering phase to ensure that accurate feature details can be learned. In the second phase, SDC fuses SIR and CIR as the final feature for clustering to interactively promote the SIR branch and CIR branch. JLIRS-UDA learns the SIR and CIR jointly in the UDA task training phase. Compared with state-of-the-arts, the strategy proposed in this paper has achieved a significant improvement of 7.1% mAP on the tasks of Market-1501 to DukeMTMC-reID. The slightest improvement in accuracy also achieved 0.8% on Market-1501 to MSMT17.
- 1
-
Fairness-aware Multi-agent Reinforcement Learning and Visual Perception For Adaptive Traffic Signal Control
Abstract:
The majority of Multi-agent Reinforcement Learning (MARL) methods for solving Adaptive Traffic Signal Control (ATSC) problems are dedicated to maximizing the throughput while ignoring fairness, resulting in a bad situation where some vehicles keep waiting. For this reason, this paper models the ATSC problem as a Partially Observable Markov Game (POMG), in which a value function that combines throughput and fairness is elaborated. On this basis, we propose a new cooperative MARL method FA-MAPPO, i.e., fairness-aware multi-agent proximity policy optimization, which is based on the cooperative MARL algorithm MAPPO. In addition, FA-MAPPO uses graph attention neural networks to efficiently extract state representations from traffic data acquired through visual perception in multi-intersection scenarios. Experimental results in Jinan and synthetic scenarios confirm that FA-MAPPO improves fairness while guaranteeing passage efficiency compared to the SOTA methods.
- 1
-
Effect of unstable resonator stability on laser beam quality
liang xiao lin, bao ben gang, zhou song qing, chen hui huang
Abstract:
In order to explore the effect of unstable resonator stability on laser beam quality, the numerical simulation of mid-infrared laser and visible laser was carried out in GLAD software. The simulation results showed that the existence of defocus aberration, tilt aberration and astigmatic aberration in the unstable resonator can cause the center of the far-field spot of the output annular beam to drift, the number of peripheral diffraction rings to increase, the beam quality to deteriorate, and the degree of effect is different. It is also found that on the basis of the effect of tilt aberration and astigmatism aberration, the introduction of defocus aberration can improve the output laser beam quality to a certain extent. In addition, under the condition of the same aberrations, the effects of different wavelength lasers are roughly the same. However, in terms of the degree of effects, the short-wave laser is much higher than the medium-long-wave laser, which verifies that the optical cavity debugging of the short-wave laser is more difficult than that of the medium-long-wave laser in the experimental process. The simulation results can provide an important reference for the optimization design of the laser system, the processing of cavity mirror and the formulation of the correction range index of the adaptive optical system.
- 1
-
Infrared Small Target Detection Based on Spatial Attention Density Peaks Searching
Abstract:
The detection of small targets poses a significant challenge for infrared search and tracking (IRST) systems, as they must strike a delicate balance between accuracy and speed. In this letter, we propose a detection algorithm based on spatial attention density peaks searching and an adaptive window selection scheme. First, the DoG filter is introduced for preprocessing of raw infrared images. Second, the image is processed by spatial attention density peaks searching (SADPS). Third, an adaptive window selection scheme is applied to obtain window templates for the target scale size. Then, the small target feature is used to enhance the target and suppress the background. Finally, the true targets are segmented through a threshold. The experimental results show that compared with the seven state-of-the-art small targets detection baseline algorithms, the proposed method not only has better detection accuracy, but also has reasonable time consumption.
- 1
-
A method to determine the complex refractive index dispersion of absorbing materials without requirement of thickness
Deng zhichao, Mei Jian Chunk, Wang Jin, Ye Qing, Tian Jian Guo
Abstract:
The complex refractive index dispersion (CRID) of absorbing materials is very important in many fields, especially in printing industry and medical research. However, due to their strong absorbing, CRID determination is still a challenge. In this study, without diluting treatment or the thickness information, a method is proposed to calculate the CRID of ab-sorbing materials, based merely on the reflectance and transmittance spectra measurements. The method separates the CRID into absorbing part and transparent part based on Kramers-Kronig relations, and it also uses the common Cauchy dispersion formula and Fresnel reflection formula. The CRID of methyl-red-doped poly (methyl methacrylate) (3% mass fraction) and hemoglobin solutions (320g/L) are determined over the spectral range from 400nm to 750 nm, and the result shows good stability and consistency of the method.
- 1
-
MZM Nonlinear Equalization by Sinusoidal Subcarrier Modulation Combined with LM-BP Neural Network
Abstract:
In order to reduce the nonlinear effect of Mach-Zehnder modulator (MZM) on optical transmission signals in intensity modulation and direct detection (IM-DD) systems, a joint scheme of sinusoidal subcarrier modulation (SSM) and Levenberg-Marquardt back propagation (LM-BP) neural network is proposed in this paper. The scheme uses sine wave as the subcarrier to carry the PAM4 signals to equalize the distorted signals after MZM modulation, and then the LM-BP algorithm removes the remaining inter-symbol interference (ISI). This scheme uses sine wave modulation to solve the problem of additional ISI caused by triangular wave modulation. And the joint scheme reduces the complexity of the algorithm than using only the neural network equalizer. In this paper, the performance of SSM-LM-BP scheme is simulated and analyzed in IM/DD system. The results show that the joint scheme outperforms the triangular wave modulation scheme as well as the neural network algorithm after transmitting 50Gbit/s PAM4 signals for 80km without relaying under the premise of dispersion compensation, and the BER can reach 10-5.
- 1
-
Simultaneous measurement of temperature and liquid refractive index based on fiber open Fabry-Pérot cavity and Bragg grating
Abstract:
A temperature and refractive index sensor based Fiber Bragg Grating (FBG) end surface cascade open Fabry-P?rot (FP) cavity has been designed and demonstrated experimentally. The open FP cavity has been fabricated on the end face of a FBG by dislocation fusion in this work, the open FP cavity could be used for refractive index sensing, the temperature is measured by cascading a FBG. The working principle of the sensor and the method of improving the sensitivity are analyzed by theoretical simulation. The refractive index sensitivity of the sensor is 1108.4nm/RIU, while the maximum fluctuation of the sensor stability experiment detection is 0.005nm, the results show that it has satisfactory characteristics. The sensor is a compact all-fiber structure, so it has potential applications in the field of temperature refractive index sensing, such as biomedical, capacitor electrolyte detection.
- 1
-
Translation of Steel Surface Defect Detection Algorithm with Fusion of Multiple Attention Detection Heads
Abstract:
This paper proposes an end-to-end defect detection algorithm model called YOLOv5-CJ to address the issues of missed detections and low accuracy in visual steel surface defect detection. By incorporating the C3_MSBlock module, which is designed based on Res2Net and contains multi-scale information, into the backbone network, the model can represent multi-scale features at a finer granularity level and increase the network's receptive field. In the detection head, the DyHead attention mechanism is introduced, which integrates scale awareness, spatial awareness, and task awareness, to help the model better cope with complex industrial scenes with large background interference, significant defect scale variations, and easily confused defect categories. Soft NMS is employed to enhance the recognition capability in overlapping regions. Experimental results demonstrate that the proposed YOLOv5-CJ model outperforms YOLOv5s, achieving an average precision (mAP0.5 and mAP0.5:O.95) improvement of 1.9% and 7.1% on the NEU-DET dataset, and 5.3% and 4.3% on the GC10-DET dataset, respectively, validating the feasibility and effectiveness of the proposed model.
- 1
-
Real-time detection of methane concentration based on TDLAS technology and 1D-WACNN
KAN Lingling, MIAO Kai, LIANG Hongwei, NIE Rui, YE Yang
Abstract:
In order to further reduce the cost of manually screening suitable second harmonic signals for curve fitting when detecting methane concentration by tunable diode laser absorption spectroscopy (TDLAS) technology, as well as the influence of certain human factors on the amplitude screening of second harmonic signals, and improve the de-tection accuracy, a one-dimensional wide atrous convolution (1D-WACNN) method for methane concentration detection is proposed, a real-time detection system based on TDLAS technology to acquire signal and Jetson Nano to process signal is built. The results show that the accuracy of this method is 99.96%. Compared with other methods, this method has high accuracy and is suitable for real-time detection of methane concentration.
- 1
-
Time Domain Characteristic Analysis of Non-coupled PCNN
Deng xiangyu, Yu haiyue, Huang xikai
Abstract:
Pulse coupled neural network(PCNN) is a multi-parameter artificial neural network, the characteristics of PCNN network can be fully explored by analyzing different simplified networks. In this paper, the firing characteristics of non-coupled PCNN model with linking terms are studied, by analyzing the firing mechanism of neurons, and the mathematical expressions of firing time and firing interval are summarized. By constructing different neighborhood linking weight matrix models, the influence of linking weight matrix and linking weight coefficient on network characteristics is further analyzed, and the constraint conditions of parameters are given. Finally, the correctness of theoretical analysis is verified through experimental simulation, and the analysis of different connection characteristics of PCNN network is further improved.
- 1
-
MAG molten pool edge detection algorithm based on a fusion of dark channel prior dehazing and image enhancement
Abstract:
Metal Active Gas welding is one of the widely applied welding techniques using argon and carbon dioxide as shielding gas. In response to the problem of welding halo and drag shadow during the image acquisition process of it, which makes it difficult to accurately extract the contour of the molten pool, this paper proposes a molten pool edge detection method that combines dark channel prior dehazing and improved single scale Retinex image enhancement algorithm. This method overcomes the problem of excessive edge noise in the original molten pool image and the difficulty in feature extraction caused by the dark part of the molten pool after dark channel prior dehazing processing. Through comparative experiments and ablation experiments, it has been shown that the algorithm proposed in this paper has significantly improved the enhancement effect and feature extraction effect, extracting accurate and complete molten pool contours.
- 1
-
Extraction of Weak Values during Reversing Quantum Entanglement State Process
Du Shaojiang, Feng Hairan, Peng Yonggang
Abstract:
A reversible operation protocol is provided for a weak-measured quantum entanglement state. The evolution of weak values is studied under different parameter conditions. The weak values can be extracted from the entanglement state and the weak-measured quantum entanglement state can be revived to its initial state theoretically by weak measurement and reversibility operation respectively. We demonstrate the reversible operation protocol by taking Bell’s state as an example. The negativity is used to analyze the initial state, the weak-measured state and the reversed state in order to describe the evolution of quantum entanglement degree. Weak values is detected from the quantum entanglement state by weak measurement and the degree of the weak-measured quantum entanglement state can be revived to its initial state through reversible operation. The information of quantum entanglement state would be extracted from weak values detected in the process of the scheme.
- 1
-
EAE-Net: Effective And Efficient x-ray joint detection
Wu Zhichao, Wan Mingxuan, Bai Haohao, Ma Jianxiong, Ma Xinlong
Abstract:
The detection and localization of bone joint regions in medical X-ray images are essential for contemporary medical diagnostics. Traditional methods rely on subjective interpretation by physicians, leading to variability and potential errors. Automated bone joint detection techniques have become feasible with advancements in general-purpose object detection. However, applying these algorithms to X-ray images faces challenges due to the domain gap. To overcome these challenges, a novel framework called EAE-Net is proposed. It incorporates a Context Augment Module (CAM) to leverage global structural information and a Ghost Bottleneck Module (GBM) to reduce redundant features. The EAE-Net model achieves exceptional detection performance, striking a balance between accuracy and speed. This advancement improves efficiency, enabling clinicians to focus on critical aspects of diagnosis and treatment.
- 1
-
Physical Layer Security of FSO Communication System based on G-G Correlation Channel
Abstract:
In this paper, based on the eavesdropping model of eavesdroppers near legitimate users, the effect of atmospheric channel correlation on the physical layer security of free-space optical link is analyzed. According to the joint probability density function and cumulative distribution function of gamma-gamma distribution, a new closed-form expression of intercep-tion probability is derived. Numerical results show that the interception probability of free-space optical system depends on turbulence intensity, channel correlation and radial displacement attenuation of eavesdroppers.
- 1
-
Broadband and polarization-independent arbitrary ratio integrated optical power splitter built on thick silicon nitride platform
Wang Linghua, Zheng langteng, Chen Yiqiang, Xue Zhengqun, Zhang Yazhen, Huang Jiwei, Minmin Zhu
Abstract:
Integrated optical power splitters are basic but indispensable on-chip devices in silicon photonics. They can be used either for power distribution or monitoring, or as the building blocks for more complex devices or circuits. Although different types of optical power splitters with different architectures have been proposed and demonstrated, devices that could work with arbitrary power splitting ratio in a large bandwidth without polarization dependence, are still rare to be seen. In this paper, we propose and investigate an optical power splitter with adiabatically tapered waveguide structures on a thick silicon nitride platform, which could meet the requirement mentioned above. With optimized structural parameters obtained by three-dimensional finite-difference time-domain simulation, the po-larization dependence of different power splitting ratio gets almost eliminated for each specific working wavelength. In a broad wavelength range (1340-1800nm), the IL of the device is below 1dB, and the variation of the PSR can be controlled within ~?5% if compared with the targeted design value for 1550nm centered wavelength. Simple structure, relaxed critical dimensions, and good fabrication tolerance make this device compatible with the standard fabrication process in commercial silicon photonic foundries.
- 1
-
Broadband and perfect terahertz absorber based on multilayer metamaterial with cross-ring patterned structures
Abstract:
Broadband and perfect terahertz absorber based on multilayer metamaterial using cross-ring patterned structures is pro-posed and investigated. The structure of the absorber is double absorption layer consisting of a chromium cross ring and eight isosceles right triangles. The unique structure of the double absorbing layers excites the electric dipole multimode resonance, giving rise to high absorption performance. Meanwhile, the influence of construal parameters on absorber be-havior is also discussed. The numerical results shows that the absorption achieves over 90% ranged from 2.45 THz to 6.25 THz and 99% absorption in the range of 3.7 THz-5.3 THz. The realization of broadband and perfect absorber is de-scribed using the impedance matching principle. It obviously found that the absorber is insensitive to the high angle of incidence for both TE and TM polarizations. Compared with the former reports, this absorber has remarkable improved absorption efficiency and smaller period. The terahertz absorber may be found applications in the fields of energy capture and thermal detection.
- 1
-
RiSw: Resistant to Incomplete Shooting Watermarking Scheme
Zhouliang Wang, Wanni Xiang, Weiya Wang, Hui Li
Abstract:
Leaking data through screen-shooting has become the main way of modern leaks. Digital watermarking technology can trace the leaker through the watermark information after the data is leaked. The current screen-shooting watermarking scheme can resist part of the distortion in the screen-shooting process, but it faces two problems. On the one hand, the watermark capacity is small. On the other hand, when the shot watermarked image is incomplete, high watermark extraction accuracy cannot be guaranteed. Based on the above problems, we propose a watermarking scheme RiSw that can resist incomplete shooting. Specifically, we design a set of codecs that can embed binary image as watermark into carrier image and extract them, which not only ensures good visual effects of watermarked image, but also greatly increases watermark capacity. To resist incomplete shooting, we propose an incomplete shooting layer to simulate the situation of incomplete shooting in the screen-shooting process. Robustness to incomplete shooting can be achieved through end-to-end training. Extensive experiments show that the scheme proposed in this paper has superior performance. Even if the watermarked image lacks 50\% pixels, it can still maintain a stable extraction accuracy.
- 1
-
A novel dual-core fiber-optic gyroscope with independ-ent rotation rate measurements in different cores of a dual-core optical fiber
Emil Anvyarovich Milikov, Viacheslav Viktorovich Zemlyakov, Pavel Sergeevich Anisimov, Jiexing Gao
Abstract:
We introduce an all-fiber stationary phase shifter for a fiber-optic gyroscope which simultaneously provides phase shifts of opposite signs in different cores of the dual-core optical fiber. We propose a new dual-core fiber-optic gyro-scope in which different cores of the dual-core optical fiber provide independent rotation rate measurements. The device enables implementation of a differential scheme, which ensures the stability of the measured phase shift. As a computer simulations result, the accuracy of the rotation rate sensing increased by up to 10 times at typical noise levels.
- 1
-
An Edge Computing-Based Embedded Traffic Information Processing Approach: Application of Deep Learning in Existing Traffic Systems
ping hao yu, ma yong jie, zhu guang ya, zhang jia qi
Abstract:
To address traffic congestion, this study improved MobileNetv2-YOLOv4 target detection algorithm (MobileNetv2-YOLOv4-K++F) and introduces an embedded traffic information processing solution based on edge computing. We transition models initially designed for large- scale GPUs to edge computing devices, maximizing the strengths of both deep learning and edge computing technologies. This approach integrates embedded devices with the current traffic system, eliminating the need for extensive equipment updates. The solution enables real-time traffic flow monitoring and license plate recognition at the edge, synchronizing instantaneously with the cloud, allowing for intelligent adjustments of traffic signals and accident forewarnings, enhancing road utilization, and facilitating traffic flow optimization. Through on-site testing using the RK3399PRO development board and the MobileNetv2-YOLOv4-K++F object detection algorithm, the upgrade costs of this approach are less than one-tenth of conventional methods. Under favorable weather conditions, the traffic flow detection accuracy reaches as high as 98%, with license plate recognition exceeding 80%.
- 1
-
Optical SSB Filters Free Twin-SSB Signal Detection with Self-Coherent Detection
Abstract:
Dater center interconnection has stimulated the development of the short reach optical communication transmission. To increase the capacity of the SSB system with DD, the twin-SSB system can double the system capacity without extra op-tical modulator. Recently, Stokes vector receiver (SVR) bridges the coherent detection and DD. A carrier can be sent along with the signal so that the SVR can utilize the reference carrier to conduct self-coherent detection and realize a linear complex optical channel detection which is similar to the coherent receiver. In this paper, we propose an optical SSB filter and guard band free twin-SSB signal detection scheme with self-coherent detection based on SVR. The proposed scheme greatly reduces the implementation complexity and has higher spectral efficiency (SE) compared with the traditional twin-SSB signal detection which two narrow-band optical filters and guard band are used for SSB signal extracting and inter crosstalk suppression. Meanwhile, the proposed twin-SSB signal detection with self-coherent detection also helps to relax the receiver DSP, since the self-coherent detection reduces the inter crosstalk between the twin-SSB. The twin-SSB reception with self-coherent detection scheme makes it suitable for the future high-speed short and medium-reach applica-tions, such as the data center interconnect and metro area network.
- 1
-
Study of PAPR mitigation in OFDM-VLC system by SCMA codebook design
yang ting, li ganggang, wang ping, wang zhao
Abstract:
Traditional orthogonal frequency division multiplexing (OFDM) based visible light communication (VLC) system is susceptible to high peak-to-average power ratio (PAPR), thus leading to low power efficiency. To address this issue, a sparse code multiple access (SCMA) codebook design method has been proposed to lower the PAPR of the clipping based OFDM-VLC system. Then, the codebooks of the high-dimension mother constellation (MC) and low-dimension MC are optimized, respectively. Specifically, for high-dimension MCs like T16-QAM, the constellation points (CLPs) with higher transmit power are mapped to the CLPs with lower power. For low-dimension MCs like 4-PAM, a contract-ed mapping method is proposed to reduce the power of the CLPs located on the real axis of the MC. Simulations show that the proposed method could achieve better PAPR performance for any values of the clipping ratio with only a slight bit error rate (BER) performance loss compared to the original codebook design. Besides, the smaller clipping ratio would induce better PAPR performance but worse BER performance. Moreover, the increasing number of iterations in the logarithm domain message passing algorithm (log-MPA) could improve the BER performance. This work will benefit the research and development of SCMA-OFDM-VLC systems.
- 1
-
Improved YOLOv5 foreign object detection for transmission lines
ZHOU Liming, LI Shixin, ZHU Zhiren, CHEN Fankai, LIU Chen, DONG Xiuhuan
Abstract:
The traditional transmission line detection has the problems of low efficiency. For improve the performance of transmission line foreign object detection, this paper proposes an improved YOLOv5 transmission line foreign object detection algorithm. First, efficient channel attention (ECA) attention module is introduced in the backbone network for focusing the target features and improving the feature extraction capability of the network. Secondly, bilinear interpolation upsampling is introduced in the neck network to improve the model detection accuracy. Finally, by integrating the EIoU loss function and Soft non-maximum suppression (Soft NMS) algorithm, the convergence speed of the model is accelerated while the detection effect of the model is enhanced. Relative to the original algorithm, the improved algorithm reduces the number of parameters by 16.4%, increases the mean average precision(mAP)@0.5 by 3.9%, mAP@0.5:0.95 by 6.3%, and increases the detection speed to 55.3FPS.The improved algorithm is able to improve the performance of the foreign object detection in transmission lines effectively.
- 1
-
MCR-YOLO model for underwater target detection based on multi-color spatial features *
Abstract:
Underwater target detection, a critical aspect of underwater robotics and ocean information processing, presents formidable challenges due to the distinctive imaging conditions beneath the water's surface. Issues such as noise interference, indistinct texture features, low contrast, and color distortion necessitate advanced computer vision techniques. In this paper, we introduce the MCR-YOLO model, a deep neural network for underwater target detection. This model leverages the Multi-Color Spatial Feature framework to address these challenges. It employs two feature extraction branches: the first focuses on the luminance channel (Y) of the RGB image in the YCbCr color space to extract non-color features, using an enhanced ResNet50 architecture. The output features from three scales are integrated for information exchange. Additionally, low-frequency information is incorporated through a separate feature extraction branch, operating on the three-channel RGB image. The features obtained from the three scales in both branches are fused at corresponding scales, enabling comprehensive feature integration. The culmination of this process results in multi-scale feature fusion and robust target detection, integrating the PANet framework. This innovative approach promises to significantly enhance the reliability and accuracy of underwater target detection in challenging underwater environments.
- 1
-
A Lightweight Semantic Segmentation Algorithm Integrating CA/ECA-Net Modules
madongmei, guozhihao, luoxiaoyun
Abstract:
Aiming at the existing semantic segmentation process due to the loss of pixel features and the complexity of calculating too many parameters, which leads to unsatisfactory segmentation results and too long time, this paper proposes a lightweight semantic segmentation algorithm based on the fusion of multiple modules. The algorithm is based on the PSPNet network, firstly, MobileNetV2 network is chosen as the feature extraction network to construct the lightweight network structure; in the training of the network, a freeze and thaw method is used, and the Focal Loss loss function is added to balance the proportion of positive and negative samples; after that, spatial and channel reconstruction convolution SCConv is introduced in the pyramid pooling module to reduce the segmentation task. The computational cost due to redundant feature extraction is reduced; finally, the coordinate attention mechanism CA and the convolutional attention mechanism ECA-Net are incorporated to make the multi-modules integrate with each other to enhance the salient features and improve the segmentation accuracy. Through the ablation and comparison experiments, the average pixel accuracy on PASCAL VOC 2012 dataset reaches 85.23%, the computation amount is reduced by 45.79%, and the training speed is improved by 68.69%; the average pixel accuracy on Cityscapes dataset reaches 86.75%, and the average intersection and merger ratio reaches 73.86%，and the interaction of multiple modules with correlation performance makes the algorithm is improved and optimised, effectively solving the problems of low segmentation accuracy and slow training speed in the algorithm, which has a significant accuracy advantage in the lightweight model, and can generally improve the efficiency of image semantic segmentation.
- 1
-
An all-dielectric chiral metasurface with circular dichroism and asymmetric transmission characteristics
Abstract:
This article designs an all-dielectric metasurface with tunable chiral properties in the near-infrared range, which working wavelength is 1250-2200 nm. The metasurface exhibits circular dichroism (CD) and asymmetric transmission (AT) characteristics for circularly polarized light. The metasurface is a double layer structure composed of Ge2Sb2Se4Te1(GSST). The CD values of amorphous GSST reach 0.83 and 0.82 at 1570 nm and 1640 nm. The AT values reach 0.65 and 0.77 at 1570 nm and 1640 nm, with a value of -0.62 at 1680 nm. The CD value of crystalline GSST reaches 0.81 at 2070 nm, with smaller AT. In addition, whether it is incident by linearly polarized waves or circularly polarized waves, this metasurface has low absorption in the working band, which gives it the potential to make adjustable integrated photonic devices.
- 1
-
EBD-YOLO： A lightweight auxiliary driving detection algorithm based on YOLOv5
Sun LianShuai, Shi Tao, Ding Yao, Li Song
Abstract:
A lightweight road-assistant detection algorithm, EBD-YOLO, based on YOLOv5s is proposed to address the problems of high model complexity, computation cost, and difficulty in deployment on resource-limited embedded terminals in existing assisted driving detection algorithms. First, lightweight Transformer model EfficientViT was used as the backbone feature extraction network of YOLOv5s model to reduce network parameters and calculation costs. Secondly, a Focal-GIoU Loss function is proposed for bounding box regression to accelerate model conver-gence and reduce loss. Thirdly, the feature pyramid structure is improved to a weighted bi-directional feature pyramid network (BiFPN) to enhance localization and semantic features. Then, a dynamic head framework is added to unify the attention mechanism with the object detection head to improve its performance. Finally, a Soft-CIoU_NMS algorithm is proposed in the post-processing stage to enhance occluded targets' localization and detection ability and reduce the missed detection rate. We conducted experiments on the KITTI and BDD100K datasets for autonomous driving, and the results showed that the EBD-YOLO model reduced in size by 38.4% and 37.2%, respectively. In comparison, the computational cost was reduced by 48.1%. As measured by mAP@0.5, the detection accuracy improved by 0.5% and 5.8%, respectively, and mAP@0.5:0.95 improved by 2.8% and 7%, re-spectively. These improvements satisfied the requirements for deployment on embedded terminals in cars.
- 1
-
Shape recognition and size measurement of particles in hybrid particle field based on interference technology
Jinlu Sun, Yue Qiu, Yuhang Wu, Dan Zhao, Changyun Miao
Abstract:
An algorithm that can implement particle shape recognition and size measurement in hybrid particle field is proposed. Based on the defocused images obtained by the interferometric particle imaging system, shape recognition of particles can be realized through ResNet50. 2D Fourier transform and 2D autocorrelation transform are used to obtained the size of spherical and non-spherical particles, respectively. The shape and size of the particle in hybrid particle field are determined. Numerical simulation and experiment results suggest that the method has good accuracy in measuring the particle size of hybrid particle field.
- 1
-
An ultra-wideband circularly polarized microstrip patch antenna loaded with split-ring resonator
Wang Shike, Xia Dan, Lv Lianrong, Qin Juan, Wang Meng, Ren Haiping
Abstract:
In this paper, the effect of a single-opening split-ring resonator (SRR) structure on the performance of ultra-wideband (UWB) microstrip patch antenna (MPA) is studied. Here a UWB MPA based on partial grounding is proposed, and the SRR structure is utilized to optimize other characteristics of the low-profile microstrip antenna while ensuring the bandwidth and gain remain unchanged. The proposed metamaterial structure, consisting of a 2 ? 2 array of SRR, is loaded on the reverse side of the substrate. The volume of the proposed antenna is 61.6?67.6?1.5mm3. According to simulation and experimental results, the proposed microstrip antenna loaded with 2?2 SRRs performs better than that without SRR structure. It produces an operating band of 1.62GHz-8GHz while achieving circular polarization characteristics with an axial ratio (AR) bandwidth of 3.92GHz-4.57GHz. In addition, the performance of the proposed antenna is superior to relevant works of literature The antenna can be widely utilized in wireless communication systems.
- 1
-
Performance Analysis of Two-Way Relay Mixed RF/FSO System with Co-channel Interference
Abstract:
In this paper, the performance of a two-way relay (TWR) mixed radio frequency/free space optical (RF/FSO) system with co-channel interference (CCI) is investigated. Different from interference-limited one-way relay (OWR) system, CCIs should be considered at both relay and users in TWR system. Additionally, opportunistic user schedule is applied to select the user with the largest signal to interference plus noise ratio (SINR). Exact and asymptotic outage probability (OP) expressions are derived. Bit error rate (BER) is further presented to evaluate the overall performance. Simulation results show that the performance of a TWR system is dominated by the uplink or downlink whose performance is severer.
- 1
-
Federated learning-based Lightweight Selective Feature Fusion and Irregular-Aware Network for Crack Detection
Liu Hui, He Tian, Cheng Xu, Shi Fan, NIKOLIĆ Saša V., ZHANG Jianhua
Abstract:
Employing deep learning for pavement crack detection can significantly enhance accuracy, and federated learning can help to overcome the challenges of data silos and data security. This paper proposes a lightweight method to detect cracks. We use lightweight encoder modules to extract multi-scale features, and further feature fusion and modelling by Selective Fusion Blocks and Irregular-aware Blocks. Moreover, this method is the first to combine federated learning with crack detection, effectively resolving the tension between data privacy and data sharing in decentralised devices. Experiments were conducted to compare our proposed method with other crack detection methods on two publicly available datasets, including the original method without federated learning and five state-of-the-art methods combined with the same feder-ated learning framework. The experiment proves that our model obtains comparable F1 score with minimal parameters and computational effort. For example, compared to the original method, the number of parameters and computation are reduced by 67.1% and 73.6%, respectively.
- 1
-
Learning to See Speckle in the Weak Laser Field through Multimode Fiber
Ji Yun Qi, Song Bin Bin, Li Xue Qing, Li Yong Hui
Abstract:
Multimode fibers (MMFs) have great potential for endoscopic imaging due to the high number of modes and a small core diameter. Deep learning based on neural networks has received increasing attention in the field of scattering image reconstruction. However, most studies focus on designing complex network architectures to improve reconstruction, but these network models struggle to reconstruct images in a weak laser field. In the paper, a lightweight generative adversarial network model combined with a histogram specification algorithm is designed to reconstruct speckles in the weak laser field through MMF. Experimental results show that the reconstruction results of our algorithm have better metrics. Moreover, the model demonstrates excellent cross-domain generalization ability with regards to the Fashion-MNIST dataset. It is worth mentioning that we found that the speckles after inactivation still retain the ability to be reconstructed, which enhances the robustness of the model.
- 1
-
High-sensitivity photonic crystal fiber sensor based on surface plasmon resonance
冯焕婷, Gao Jiachen, Ming Xianbing
Abstract:
In this paper, we propose a photonic crystal fiber sensor based on the surface plasmonic resonance effect for simultaneous temperature and refractive index measurement. The coupling characteristics and sensing performance of the sensor are analyzed using the full vector finite element method. The sensor provides two channels for in-dependent measurement of refractive index and temperature. When operating independently, channel I supports y-polarized light with a sensitivity of up to 7000 nm/RIU for detecting refractive index, while channel II supports x-polarized light with a sensitivity of up to 16 nm/°C for detecting temperature. Additionally, we investigated the influence of gold layer thickness on the sensing performance to optimize the sensor's performance.
- 1
-
Subnetwork-based Federated Few-shot semantic seg-mentation of Organ Images
Wu Junpeng, Zhao Meng, Zhang Huanping
Abstract:
Federated learning, as a distributed learning paradigm, allows multiple medical institutions to collaborate on learning without the need to centralize all client data. However, existing methods pay little attention to more challenging medical image semantic segmentation tasks, especially in the scenario of uneven data category distribution in fed-erated few-shot learning. In this context, we propose a subnetwork-based federated few-shot medical image seg-mentation method. Firstly, individual clients train using local training samples and then upload local model gradients to the server. The server utilizes their respective local model gradients to update the subnetwork maintained on the server and generate aggregation weights for forming personalized model parameters. Through this method, we are able to learn the similarities between different clients to address data heterogeneity issues. In addition, to enhance the communication efficiency between clients and the server, we have also designed a personalized layer aggregation strategy, which only transmits partial layer model parameters during the communication process to improve com-munication efficiency. Finally, we conducted experiments on ABD-MRI and ABD-CT datasets to demonstrate the effectiveness of our method.
- 1
-
Effect of ammonia concentration of complexing agent on CdxZn1-xS buffer layer by chemical bath deposition and performance of CIGS cell
yahan Lu, yuming Xue, hongli Dai, luoxin Wang, liwei Zhou, zhiyuan Tan
Abstract:
In this paper, CdZnS thin films were deposited on glass substrates by chemical bath deposition (CBD), and the effects of different concentrations of ammonia water on the morphology, structure and optical properties of the films were studied. CdZnS thin films have hexagonal crystal structure, the transmittance is above 75% in the visible range, and the optical band gap is between 2.6 eV and 2.9 eV. The influence of the experimental group of deposited thin films on the performance parameters of the CIGS battery was studied by simulation with SCAPS-1D software and the cell efficiency reached 22.42%. It is introduced that the buffer layer prepared in this experiment is feasible in CIGS solar cells.
- 1
-
Improved model for defect detection in photovoltaic cells based on Yolov5
baixinfeng, xueyuming, daihongli, wangluoxin, baixin, huxiaofeng
Abstract:
In order to meet the requirements of recognition speed and accuracy for complex photovoltaic (PV) cell defect detection tasks, this paper proposed an improvement strategy based on the Yolov5 network model. First, the model's ability to localize defects was enhanced by adding shallow feature inputs to strengthen the model's fusion of shallow feature information; second, the Squeeze-and-excitation (SE) attention mechanism was introduced to expand the receptive field and strengthen the dependency between feature channels; finally, a smoothing downsampling module was constructed to replace the convolutional sampling, which reduced the loss of information while retaining the important features. The experimental results show that the Yolov5 model based on this improved strategy achieved an average accuracy of 92.06% in the multiclassification defect detection task, and the average recognition speed reached 25.31 frames per second (FPS), which took into account the requirements of performance and real-time performance, and can satisfy the needs of PV cell defect detection in industrial scenarios.
- 1
-
Study on the properties of solar cell under bias condition by using impedance spectroscopy
Abstract:
This article investigates the properties of silicon (SI) and perovskite (PSC) solar cells under bias condition by using impedance spectroscopy. The parallel resistance Rp of SI and PSC are found to decrease with increasing bias, but the capacitance Cp shows the opposite trend. Comparing Rp to Cp, bias has a greater impact on the Cp of both cells. However, the bias has little effect on the series resistance Rs of both cells. Compared with SI cell, the bias is seen to be a greater impact on the Rp and a smaller impact on Cp of PSC. The carrier lifetime in SI and PSC cells first increases and then decreases with bias, and the carrier diffusion length Ld increases almost linearly with bias. Compared to SI cell, the bias has a greater impact on PSC’s Ld.
- 1
-
Research on φ-OTDR signal pre-filtering based on spa-tial similarity
Abstract:
A theoretical analysis was conducted on the intrinsic bond between multi-point response caused by a same single vibration in phase sensitive optical time domain reflectometer. Temporal similarity of signals collected from adja-cent sample locations were investigated. Referring to correlation coefficient as well as the relative energy level, a method of extracting disturbed position in φ-OTDR based on signal relevance evaluation is proposed to perform fast screening of massive φ-OTDR raw data to pinpoint those signals with significance . As proof of concept, a manual excavation experiment was conducted along an oil pipeline, where on-site data was analyzed. The results showed that the proposed method can accurately screen out real vibration signals and filter out pure noise so that compu-tation resources could be allocated with better rationality.
- 1
-
High-quality environmentally friendly photovoltaic CZTSSe thin films prepared by Mg doping
Zhang Zihe, Xue Yuming, Dai Hongli, Wang Luoxin, Bai Xin, Hu Xiaofeng
Abstract:
CZTSSe thin films were prepared using the sol-gel method, and the crystal morphology of the CZTSSe films were improved by Mg doping. The prepared films were characterized using techniques such as XRD, Raman spectroscopy, SEM, and UV-Vis spectroscopy. The results showed that Mg replaced Zn in the CZTSSe lattice, forming the CMZTSSe phase. As the Mg doping concentration increased, the grain size initially increased and then decreased.After Mg doping, no additional impurities are produced.When the Mg doping concentration was 0.1, the film exhibited the optimal crystal morphology, narrowest peak width, largest grain size, the best light absorption propertiessmoothest and most compact surface, which is favorable for use as an absorber layer in solar cells.
- 1
-
Double Random Number Encryption Blind Watermarking Technique Based on DWT-DCT Domain
Abstract:
Blind watermarking plays a significant role in copyright protection and information hiding. To enhance the security of watermark image information, a dual encryption method is proposed, which combines block shuffling with logistic chaos encryption using two sets of random numbers as keys. This dual encryption technique is applied to blind wa-termark images, utilizing a combination of discrete wavelet transform (DWT) and discrete cosine transform (DCT) as the embedding method. First, the watermark image is divided into small blocks, which are shuffled and rearranged using one set of random numbers as the key, while recording the shuffling sequence of block positions. Subsequently, the shuffled blocks are reassembled into an image, and another set of random numbers is used as the key for logistic chaos encryption, rendering the watermark image chaotically invisible. Then, a double discrete wavelet transform (DWT) is applied to the carrier image, followed by a discrete cosine transform (DCT), allowing the encrypted water-mark image to be embedded. This achieves the concealment of the blind watermark. Experimental results demonstrate that this method exhibits robustness, excellent invisibility, and secrecy.
- 1
-
A rendering method for predicting sampling distribution based on lighting and JND information
Abstract:
When using Monte Carlo path tracing method to render 3D scenes, artifacts may occur due to insufficient sampling. Di-rectly increasing the number of samples can increase the time cost of the rendering algorithm. An effective strategy is to raise sampling levels iteratively. However, iterative operations introduce additional computational overhead. To solve this problem, we proposed a rendering acceleration method that does not require iterative computations. This method com-bined extraction of the Just Noticeable Difference (JND) information and used a neural network to predict the sampling matrix of the scene, which was adjusted based on the lighting information in the pre-rendered image. To start with, we extracted the JND information from pre-rendered images and estimated the rapid convergence regions, such as environ-ment mapping regions and light source regions. We then employed the Conv-LSTM to estimate the JND features for high-quality rendered images. We then employed the Conv-LSTM to estimate the JND features for high-quality rendered images. We designed a multi-feature fusion network to predict the required number of samples for each pixel. The en-coder took the pre-rendered images as input, which were then fused with the JND features for the decoder to generate the corresponding sampling matrix. In addition, based on the observation about the slow convergence at low lighting areas, we adjusted the sampling matrix according to the lighting clustering results derived from the pre-rendered images. The experimental results indicated that our method has better performance compared with the current methods.
- 1
-
Study on impedance spectroscopy based on dynamic equivalent circuit of solar cell
wenbo xiao, Li Ao, Wu Huaming, Li Yongbo
Abstract:
This paper investigates the impedance spectroscopy of monocrystalline silicon solar cells (MSSC) and dye-sensitized solar cells (DSSC) using solar cell dynamic equivalent RC and CSC circuits. Firstly, these circuits effectively represent the dynamic behavior of MSSC and DSSC. Secondly, the measurement method significantly impacts the accuracy of impedance measurements in the high-frequency region. Finally, the series resistance affects the distance between the left end of the impedance spectroscopy and the origin, while the parallel resistance influences the size of the impedance spectroscopy. In the CSC circuit, the relative magnitudes of dielectric relaxation capacitance and chemical capacitance affect the number and position of arc in the impedance spectroscopy. The value of the dielectric relaxation capacitor determines the number of impedance spectroscopy arc. These conclusions provide guidelines for improving the accuracy of solar cell impedance measurements.
- 1
-
A fast Streptomycin Sulfate Sensing based on Nb2CTX functionalized titled fiber Bragg grating*
Xiaolan Li, BI Mingpan, LIU Jie, ZHANG Yuchen, MIAO Yinping
Abstract:
We proposed a fast and temperature self-compensation method to detect streptomycin sulfate aqueous solution concentration based on a 2D Nb2CTX MXene nanosheets functionalized tilted Bragg grating(N-TFBG). The in-troduction of 2D Nb2CTX MXene nanosheets brought more than 100% increase in sensitivity and exhibited a high sensitivity of 253.8 pm/mg/mL and detection limit 80 µg/mL in range of 0.2-0.8 mg/mL. Due to high sensi-tivity and compact structure, N-TFBG sensor has a faster measure time than previous method; additionally, the sensor has simultaneous temperature measurement by detecting the cladding mode and the core mode resonance peak simultaneously. Our research will hopefully open a new research platform for online detection of strepto-mycin sulfate concentration.
- 1
-
Investigation On FBG Based Optical Sensor For Pressure And Temperature Measurement In Civil Application
Somesh, Dr Chethana K, Dr. T Srinivas, Dr Preeta Sharan
Abstract:
Optical fibre Bragg grating (FBG) sensors have advanced significantly in the last several years. The use of innovative FBG in temperature and pressure measurement is examined in this study, the benefits of FBGs, such as their compact size, low weight, resilience to corrosion, immunity to electromagnetic interference, distributed sensing, and remote monitoring, have brought attention to the growing research in this field of structural health monitoring of civil infrastructures. In this investigation a novel model is proposed and implemented using ANSYS workbench and Grating MOD tool , it was shown that the centre wavelength of FBG sensors increased from 1550nm to 1556nm when the temperature rose from 10°C to 40°C. In a similar vein, the centre wavelength grew from 1551.1667 nm to 1560.056 nm over a pressure range of 100 MPa to 600 MPa. The claimed work will make it possible to calibrate sensors more precisely, guaranteeing accurate data and being useful in monitoring numerous parameters at once, making them beneficial in a variety of applications.
- 1
-
Design of high quality-factor Si3N4 ridge-slot micro-ring resonator and generation of dual-comb based on orthogonally bicolor pumping
Wen Jin, Wang Qian, Yu Huimin, Wu Zhengwei, Zhang Hui
Abstract:
A novel high quality-factor (Q) micro-ring resonator (MRR) structure based on the Si3N4 ridge-slot waveguide is proposed, and the MRR is pumped by orthogonally polarized bicolor pumping to generate dual-comb. We optimized the structure of MRR by the finite element method and precise dispersion engineering, which finally obtained the suitable MRR geometry with negative dispersion characteristics at 1550 nm, having Q of 1.7?10^7 and the absorption loss as low as 2.6?10^-5 dB/cm. The simulation model of generating dual-comb is established as coupled Lugiato-Lefever equation (LLE), which takes into account the higher order dispersion, cross-phase modulation (XPM), multiphoton absorption, and external pumping. Solved by the split-step Fourier method (SSFM) and the fourth-order Runge-Kutta (RK4) method, the numerical results show that the generated dual-comb is periodically equally spaced distribution but with slightly different intensities in the time domain. In the frequency domain, there are 64 comb teeth with intensities higher than -100 dBm with a bandwidth of 120 nm. Particularly, in the case of bicolor orthogo-nal polarization pumping, a smaller amount of detuning does not greatly affect the bandwidth of the dual-comb.
- 1
-
Simulation of random fiber Bragg grating array in polarization-maintaining fiber based on photonic localization effect
Abstract:
Mid-infrared wavelength switchable and dual-wavelength random laser (RL) output has many potential applications. A polarization-maintaining random fiber Bragg grating (PMRFBG) array based on the photonic localization effect of longitudinal invariant transverse disorder in fiber structure is proposed, which can be used as random feedback of dual-wavelength and wavelength switchable output of random fiber laser. The random fiber Bragg grating (RFBG) array was de-signed on the Panda-type polarization-maintaining fiber (PMF), and the two center wavelengths were 2151.60 nm and 2152.22 nm, respectively; The RFBG array was designed on the bow tie-type PMF, and the two center wavelengths were obtained, which were 2153.08 nm and 2153.96 nm, respectively; The RFBG array with a center wavelength of 2139.27nm was designed on single-mode fiber (SMF). The length of individual fiber Bragging grating (FBG) and PMFBG, the refractive index modulation depth, the number of cascaded gratings, and the distance between gratings have different effects on the FWHM and reflectance of the RFBG and PMRFBG array, but not on the central wavelength, as obtained by simulation using the transmission matrix method. The designed PMRFBG array provides theoretical support for the design of the feedback mechanism of random fiber laser (RFL).
- 1
-
Defect detection of light guide plate based on improved YOLOv5 Networks
XIAO Ming, GONG Yefei, WANG Hongding, LU Mingli, GAO Hua
Abstract:
Light guide plate (LGP) is a kind of material used in backlight module. How to improve the quality control of LGP has become the focus of research in the industry. To address issues such as low gray contrast and large proportion of small target defects in LGP images, an improved YOLOv5 neural network based on multi-scale dilation convolution and a novel loss function is proposed. First, the LGP image is preprocessed, and then the Context Amplification Module (CAM) is integrated in the feature fusion part of the detection algorithm to fuse multi-scale expansion convolution features to obtain rich context information. XIoU Loss function is selected as the location regression Loss function. The result shows that this method can effectively improve the detection accuracy and positioning accuracy. Compared with YOLOv5, the average accuracy is increased by 4.7%, and the recall rate is increased by 27%. It can achieve accurate detection of defects such as white/bright spots, black spots,line scratches,and surface foreign objects in LGP.
- 1
-
High Sensitivity Dual Polarization Four Channels Surface Plasmon Resonance Refractive Index Sensor with Wide Detection Range and High Linearity Based on Photonic Crystal Fiber
cuinan, panghonggang, zhangailing, zhaozihong, chenzhipan, zhangbo, linsihang, caoguangxiao
Abstract:
In this paper, we propose and numerically characterize the optical characteristics of a high sensitivity dual polarization four channels surface plasmon resonance refractive index sensor with wide detection range and high linearity based on photonic crystal fiber (PCF). Four light transmission channels are constructed by using big air holes. The small rectangular air hole in the center of the fiber core is used to add asymmetry. We use gold plasma material to achieve better sensing performance. A thin film of TiO2 is placed on the PCF surface, which assists gold in the adhesion on PCF. The maximum wavelength sensitivity of the sensor can reach 20000 nm/RIU in both x and y polarization modes. While the maximum wavelength resolution of the sensor is 5×10-6 RIU. The average wavelength sensitivity reached 15900nm/RIU and 15800nm/RIU in x and y polarization mode. The sensor features R2=0.9969 / R2=0.9966 high linear performance in x and y polarization mode and the maximum FOM values 509.13 RIU-1 / 489.63 RIU-1 in x and y polarization mode. Besides the sensor’s detection range is 1.26-1.36, which is a wide refractive index detection range. The sensor has a wide range of applications in biomedical and other fields.
- 1
-
Effects of mixed aerosol on the path loss of NLOS UV communication system
Abstract:
Performance of non-line-of-sight (NLOS) ultraviolet (UV) communication is closely related with the communication range, system geometry and the atmosphere aerosol properties. In the paper we for the first time investigate the path loss of the NLOS UV communication systems in both monodisperse and polydisperse aerosol systems based on the Monte-Carlo method. The mixed aerosols composed of black carbon and sulfate are considered as the transmission media. The core-shell, homogeneous, and external mixing model are assumed. Simulation results show that the performance of the NLOS UV communication in the monodisperse aerosol is similar to that in polydisperse aerosol. The mixed state of the aerosols has significant influences on the performance of the communication system. The path loss of the communication system in externally mixed aerosol is smaller than that in internally mixed aerosol. Our simulation results are useful for the design of the NLOS UV communication systems.
- 1
-
Hyperbolic Cosine Transformer for LiDAR 3D Object Detection
Abstract:
Recently, Transformer has achieved great success in computer vision. However, it is constrained because the spatial and temporal complexity grows quadratically with the number of large points in 3D object detection applications. Previous point-wise methods are suffering from time consumption and limited receptive fields to capture information among points. To address these limitations, we propose the cosh-attention, which reduces the computation complexity of space and time from the quadratic order to linear order with respect to the number of points. In the cosh-attention, the traditional softmax operator is replaced by non-negative ReLU activation and hyperbolic-cosine-based operator with re-weighting mechanism. Then based on the cosh-attention, we present a two-stage hyperbolic cosine transformer (ChTR3D) for 3D object detection from point clouds. It refines proposals by applying cosh-attention in linear computation complexity to encode rich contextual relationships among points. Extensive experiments on the widely used KITTI dataset demonstrate that, compared with vanilla attention, the cosh-attention significantly improves the inference speed with competitive performance. Experiment results show that, among two-stage state-of-the-art methods using point-level features to refine proposals, the proposed ChTR3D is the fastest one.
- 1
-
Temperature-insensitive micro-displacement sensor with spindle-shaped structure single mode fiber
Weihua Zhang, Jinlin Mu, Zhengrong Tong, Haozheng Yu, Yipeng Tao, Hao Wang
Abstract:
A temperature-insensitive micro-displacement sensor based on a spindle-shaped single mode fiber (SMF) is presented and demonstrated. The SMF is bent into a balloon-shaped SMF and burned into a spindle-shaped SMF by a flame. Due to the bending of the fiber, part of the incident light leaks from the fiber core into the fiber cladding, which excites higher order cladding modes. Therefore, modal interference occurs. The experimental results show that the maximum sensitivity of the sensors is -203 pm/μm when micro-displacement varies from 0 to 80 μm. The sensors are insensitive to temperature in the range of 20-70℃. In addition, the sensors have advantages of simple structure, low cost, small volume, high sensitivity and stability, which can be widely used in numerous high-precision industrial measurements and civil engineering structure monitoring fields.
- 1
-
2SWUNet: Small Window SWinUNet based on Tans-former for Building Extraction from High-Resolution Remote Sensing Images
yujiamin, chansixian, leiyanjing, wuwei, wangyuan, zhouxiaolong
Abstract:
Models dedicated to building long-range dependencies often exhibit degraded performance when transferred to remote sensing images. Vision Transformers (ViTs) is a new paradigm in computer vision that uses multi-head self-attention ra-ther than convolution as the main computational module, with global modeling capabilities. How-ever, its performance on small datasets is usually far inferior to that of convolutional neural networks (CNNs). In this work, we propose a Small Window SWinUNet (2SWUNet) for building extraction from high-resolution remote sensing images. Firstly, 2SWUNet is trained based on Swin Transformer by designing a fully symmetric encod-er-decoder U-shaped architecture. Secondly, to construct a reasonable U-shaped architecture for building extraction from high-resolution remote sensing images, the different forms of patch expansion are explored to simulate up-sampling operations and recover feature map resolution. Then, the small window-based multi-head self-attention (W-MSA) is designed to reduce the computational and memory burden, which is more appropriate for the features of remote sensing images. Meanwhile, the pre-training mechanism is advanced to make up for the lack of decoder parameters. Finally, comparison experiments with other mainstream CNNs and ViTs validate the superiority of the proposed model. In addition, by visualizing the effective receptive field, we dis-cover that the local information is more conducive to predicting in remote sensing images.
- 1
-
Enhancing the Performance of the Absorber of Cu2ZnSn(S, Se)4 Photovoltaic Materials by Doping Cd2
shuqi yang, yuming xue, hongli dai, luoxin wang, haiwei zhang, xin bai
Abstract:
Cationic doping is considered to be an effective way to improve the efficiency of Cu2ZnSn (S, Se) 4 (CZTSSe) photovoltaic materials. In order to improve the quality of the absorption layer of the crystal film, CZTSSe film was doped with some Cd to change the ratio of Cd and Zn in the precursor solution. XRD, Raman, SEM, UV-Vis spectroscopy, and other results showed that Cd successfully replaced Zn in the crystal lattice during the post-selenization process. The formation of harmful Zn-related defects was reduced, the grain size and crystallinity of the films were significantly increased, and the photoelectric properties of the films such as crystal quality and structural morphology were improved. As the ratio of Cd/(Cd Zn) increased from 0 to 0.35, the band gap of CZCTSSe decreased to 1.02eV. When Cd/(Cd Zn)=0.25, the crystallinity and grain size reached the best value, and the film surface was smooth and dense, which inhibited the formation of the second phase on the surface of CZCTSSe. The theoretical basis and experimental results showed that proper Cd doping can promote the growth of grain, so that compact CCZTSSe films with larger grain size and fewer holes could be prepared.
- 1
-
YOLOV5s object detection based on Sim SPPF hybrid pooling
Dong xiuhuan, Li shixin, Zhang jixiang
Abstract:
Aiming at the problem of low surface defect detection accuracy of industrial products, an object detection method based on Sim SPPF hybrid pooling improved YOLOV5s model is proposed. The algorithm introduces Channel Attention module, Simplified Spatial Pyramid Pooling Fast Feature Vector Pyramid and Efficient Intersection Over Union loss function. Feature vector pyramids fuse high-dimensional and low-dimensional features, which makes semantic information richer. The channel attention mechanism performs maximum pooling and average pooling operations on the feature map. Hybrid pooling comprehensively improves detection computing efficiency and accurate deployment ability. The results show that the improved YOLOV5s model is better than the original YOLOV5s model, the average test accuracy (mAP) can reach 91.8%, the average accuracy (mAP) of the model can be increased by 17.4%, the detection speed can reach 108 FPS, the detection speed can be increased by 18 FPS, and the improved model is practicable and the overall performance is better than other conventional models.
- 1
-
MFCC based real-time speech reproduction and recognition using distributed acoustic sensing technology.
Zhou Ran, Zhao Shuai, Luo Mingming, Meng Xin, Ma Jie, Liu Jianfei
Abstract:
The distributed acoustic sensing technology was used for real-time speech reproduction and recognition, in which the voiceprint can be extracted by the Mel Frequency Cepstral Coefficients (MFCC) method. A classic ancient Chinese poem “You Zi Yin”, also called “A Traveler’s Song”, was analyzed both in time and frequency domain, where its real-time reproduction was achieved with a 116.91 ms time-delay. The smaller scaled MFCC0 at 1/12 of MFCC matrix was taken as a feature vector of each line against the ambient noise, which provides a recognition method via cross-correlation among the six original and recovered verse pairs. The averaged cross-correlation coefficient of the matching pairs is calculated 0.5806 higher than 0.1883 of the nonmatched pairs, promising an accurate and fast method for real-time speech reproduction and recognition over a passive optical fiber.
- 1
-
Composite-Mask GAN based on refined optical flow and disparity map for SLAM Visual Odometry
JI Yuehui, JIANG Jingwei, LIU Junjie, Song Yu, GAO Qiang
Abstract:
Although deep learning methods have been widely applied to slam visual odometry over the past decade with impres-sive improvements, their accuracy remains limited in complex dynamic environments. In this paper, we use a compo-site mask-based generative adversarial network to predict camera motion and binocular depth maps. Specifically, a perceptual generator is first designed to obtain the corresponding parallax map and optical flow from between two neighboring frames. Then, an iterative pose improvement strategy is proposed to improve the accuracy of pose estima-tion. Finally, a composite mask is embedded in the discriminator to sense structural deformations in the synthetic vir-tual image, thus encouraging the generator to learn additional structural level information to improve the accuracy of pose estimation. Detailed quantitative and qualitative evaluations on the KITTI dataset show that the proposed framework outperforms existing conventional, supervised learning and unsupervised depth VO methods, providing better results in both pose estimation and depth estimation.
- 1
-
Learning Background Restoration and Local Sparse Dictionary for Infrared Small Target Detection
Yue He, Rui Zhang, Chunmei Xi, Hu Zhu
Abstract:
This paper proposes a method for learning background restoration for infrared small target detection, employing a local sparse dictionary alongside an equalized structural texture representation. The method is specifically designed for the detection of small infrared targets, accommodating various levels of brightness, spatial size, and intensity. Our proposed model intelligently combines global low-rankness and local sparsity to estimate the rank of the background tensor, leveraging spatial and structural information to overcome the limitations posed by insufficient detailed texture knowledge. Subsequently, a structural texture representation, combining local gradient maps and local intensity maps, is applied to emphasize small objects. Experimental results demonstrate the clear superiority of our proposed method over existing state-of-the-art techniques.
- 1
-
A high-performance multi-wavelength optical switch based on multiple Fano resonances in an all-dielectric metastructure
Cao Shuangshuang, Fan Xinye, Fang Wenjing, Chen Huawei, Bai Chenglin, Tong Cunzhu
Abstract:
The multi-wavelength optical switch based on an all-dielectric metastructure consisting of four asymmetric semi-circular rings was designed and analyzed in this paper. Four Fano resonance modes, which can be explained by bound states in the continumm (BIC) theory, are excited in our structure with maximum Q-factor of about 2450 and modulation depth close to 100%. By changing the polarization direction of the incident light, the transmission amplitude of Fano resonances can get effectively modulated. Based on this tuning property, the metastructure can achieve a multi-wavelength optical switch and the maximum extinction ratio can reach 38.3 dB. In addition, the results indicate that the Fano resonances are sensitive to the changes of refractive index. The sensitivity(S) and the figure of merit (FOM) are 197 nm/RIU and 492 RIU-1. The proposed metastructure has promising potential in applications such as optical switches, sensors, modulators and lasers.
- 1
-
Enhancing Hyperspectral Power Transmission Line Defect and Hazard Identification with an Improved YOLO-based Model
WANG MENG, SUN LONG, JIANG JIONG, YANG JINSONG, ZHANG XINGRU
Abstract:
To address the challenges of inefficient manual inspections and time-consuming video monitoring for power transmission lines, this paper presents an innovative solution. It combines deep learning algorithms with visible light remote sensing images to detect defects and hazards. Deep learning offers enhanced robustness, significantly improving efficiency and accuracy. The study utilizes YOLO-V7 as a foundational framework, enhancing it with the Transformer algorithm, Triplet Attention mechanism, and SIoU loss function. Experimental results show a remarkable 92.3% accuracy and an 18.4 ms inference speed. This approach promises to revolutionize power transmission line maintenance, offering real-time, high-precision defect and hazard identification.
- 1
-
PointNetV3: feature extraction with position encoding
JunWang, Wang Xuefei, Zhou Boxiong, Guo Dongyan
Abstract:
Feature extraction of point clouds is a fundamental component of 3D vision tasks. While existing feature extraction net-works primarily focus on enhancing the geometric perception abilities of networks and overlook the crucial role played by coordinates. For instance, though two airplane wings share the same shape, it demands distinct feature representations due to their differing positions. In this paper, we introduce a novel module called Position-Aware Module (PAM) to lev-erages the coordinate features of points for positional encoding, and integrating this encoding into the feature extraction network to provide essential positional context. Furthermore, we embed PAM into the PointNet framework, and de-sign a novel feature extraction network, named PointNetV3. To validate the effectiveness of PointNetV3, we conducted comprehensive experiments including classification, object tracking and object detection on point cloud. The results of remarkable improvement in three tasks demonstrate the exceptional performance achieved by PointNetV3 in point cloud processing.
- 1
-
Research on Visual Sensitivity Characteristics of Amorphous Silicon Photocells
Wang yan, Cai nuo, Fan xin min
Abstract:
This study delves into the feasibility of using amorphous silicon photocells as photosensitive units for retinal pros-theses. Firstly, theoretical simulations coupled with experimental results demonstrated its strong light absorption and quantum efficiency within the 300-800 nm range. Subsequently, measurements on its visual sensitivity properties were conducted. The findings revealed that, under photopic vision conditions, the photocells could provide the stim-ulating current required for the human retinal nerve cells. Finally, the visual spectral sensitivity curve of the amor-phous silicon photocells was assessed, and the results indicated that the spectral sensitivity curve of the amorphous silicon photocells closely mirrors the visual function curve of the human eye under photopic conditions, demonstrat-ing a response to light across various wavelengths.
- 1
-
The effects of selenization conditions on the microstructure and optoelectronic properties of CZTSSe ab-sorber layers prepared by the sol-gel method
baixin, xueyuming, daihongli, wangluoxin, baixinfeng, huxiaofeng
Abstract:
CZTSSe thin film solar cells, with adjustable bandgap and rich elemental content, hold promise in next-gen photovoltaics. Crystalline quality is pivotal for efficient light absorption and carrier transport. During the post-selenization process, understanding crystal growth mechanisms, and improving layer quality are essential. We explored the effects of ramp rate and annealing temperature on CZTSSe films, using XRD, Raman, SEM, and UV-Vis. The optimal performance occurred at 25.25 ?C/min ramp rate and 530?C annealing. This led to smoother surfaces, higher density, and larger grains. This condition produced a single-layer structure with large grains, no secondary phases, and a 1.14 eV bandgap, making it promising for photovoltaic applications. The study has highlighted the effect of selenization conditions on the characteristics of the CZTSSe absorber layer and has provided valuable information for developing CZTSSe thin film solar cells.
- 1
-
Weak feedback self-mixing interference fringe slope discrimination method based on deep learning
ZHAO Yan, LIN Maohua, DU Shengzhi, TONG Jigang, LIU Bin, HAN Fangfang
Abstract:
In order to identify the tilt direction of the self-mixing signals under weak feedback regime interfered by noise, a deep learning method is proposed. The one-dimensional U-Net(1D U-Net) neural network can identify the direction of the self-mixing fringes accurately and quickly. In the process of measurement, the measurement signal can be normalized and then the neural network can be used to discriminate the direction. Simulation and experimental results show that the proposed method is suitable for self-mixing interference signals with noise in the whole weak feedback regime, and can maintain a high discrimination accuracy for signals interfered by 5dB large noise. Combined with fringe counting method, accurate and rapid displacement reconstruction can be realized.
- 1
-
Simulation and Implementation of a Reconfigurable Dual-Function Pixel
Shaher Dwik, Gurusamy Sasikala
Abstract:
This paper presents the simulation and implementation of a reconfigurable pixel that serves both data acquisition and energy harvesting purposes. The main topic focuses on switching between the two operating modes of the photo-diode: photoconductive and photovoltaic. This proposed model can be used to design novel optical sensors with energy harvesting capability, such as Position Sensitive Device (PSD) and CMOS image sensors, which can extend the battery lifetime of the whole optical system. Thus, we can overcome power supply problems like wiring and changing batteries frequently, especially in hard-to-reach places like space (cube satellites) or even Underwater Wireless Optical Communication (UWOC). The proposed pixel architecture offers the advantage of a minimalistic design with only four transistors. Nevertheless, it does come with a drawback in the form of higher noise levels. The simulation was achieved using MATLAB, and the implementation was performed using the Programmable Sys-tem-on-Chip (PSoC) microcontroller. The results showed that the functionality of the dual-function pixel is correct, and the scheduling of both energy harvesting and signal sensing functions was successfully achieved.
- 1
-
RepColor: Deep coloring algorithm combining semantic categories
Abstract:
Image coloring is an inherently uncertain and multimodal problem. By inputting a grayscale image into a coloring network, visually plausible colored photos can be generated. Conventional methods primarily rely on semantic information for image colorization. Although effective in coloring images with clear semantic information, these methods still suffer from color contamination and semantic confusion. This is largely due to the limited capacity of convolutional neural networks to effectively learn deep semantic information inherent in images.In this paper, we propose a network structure that addresses these limitations by leveraging multi-level semantic information classification and fusion. Additionally, we introduce a global semantic fusion network to combat the issues of color contamination. The proposed coloring encoder accurately extracts object-level semantic information from images.To further enhance visual plausibility, we employ a self-supervised adversarial training method. We train the network structure on various datasets with varying amounts of data and evaluate its performance using the ImageNet validation set and COCO validation set. Experimental results demonstrate that our proposed RepColor can generate more realistic images compared to previous approaches, showcasing its high generalization ability.
- 1
-
Implementation of all-optical tristate Pauli X, Y and Z gates based on two-dimensional photonic crystal.
Snigdha Hazra, Sourangshu Mukhopadhyay
Abstract:
In this paper, we have designed and simulated all-optical tristate Pauli X, Y and Z gates using 2D photonic crystal. Simple line and point defects have been used to design the structure. The performance of the structure has been analyzed and investigated by plane wave expansion and finite difference time domain methods. Different performance parameters namely contrast ratio, rise time, fall time, delay time, response time, bit rate etc. have been calculated. The main advantage of the proposed design is that all the Pauli gates have been realized from a single structure. Due to compact size, fast response time, good contrast ratio and high bit rate, the proposed structure can be highly useful for optical computing, data processing and optical integrated circuits.
- 1
-
Double-branch forgery image detection based on multi-scale feature fusion
zhanghongying, guochunxing, wangxuyong
Abstract:
Most of existing methods exhibit poor performance in detecting forged images due to the small size of tampered areas and the limited pixel difference between untampered and tampered regions. To alleviate the above problem, a dou-ble-branch tampered image detection based on multi-scale features is proposed. Firstly, we introduce a fusion module based on attention mechanism in the first branch to enhance the network's sensitivity towards tampered regions. Secondly, we construct a second branch specifically designed for detection, aiming to identify subtle differences between tampered and untampered areas by utilizing rich edge information from shallow features as guidance. Compared to the existing methods on the public benchmark datasets CASIA1.0, Columbia and NIST16, the values of F1-Score reached 0.766,0.9 and 0.93 on the those datasets respectively. The experimental results show that our method could significantly improve the accuracy on detecting the tampered area.
- 1
-
Advancements and Applications of Position-Sensitive Detector (PSD): A Review
Shaher Dwik, Gurusamy Sasikala, Somasundaram Natarajan
Abstract:
This paper presents a review of the Position-Sensitive Detector (PSD) sensor, covering the different types of PSD and recent works related to this field. Furthermore, it explained the theoretical concepts and provided information about its structure and principle of operation. Moreover, it includes the main information about the available commercial PSDs by Hamamatsu, along with a comparison between the common modules. The PSD features include high position resolution, fast response, and a wide dynamic range. These features make it suitable for various fields and applications such as imaging, spectrometry, spectroscopy and others.
- 1
-
Combining TDLAS and multi-fusion algorithms for me-thane gas concentration detectionn?
Abstract:
High-precision methane gas detection is of great importance in industrial safety, energy production and environmental protection, etc. However, in the existing measurement techniques, the methane gas concentration information is susceptible to noise, which leads to its useful signal being drowned by noise. A fusion algorithm of variational modal decomposition (VMD) and improved wavelet threshold filtering is proposed, which is used in combination with tunable diode laser absorption spectroscopy (TDLAS) to implement a non-contact, high-resolution methane gas concentration detection. The fusion algorithm can perform noise reduction and further segmentation of the methane gas detection signal. And the simulation and experiment verify the effectiveness of the fusion algorithm, and the experimental results show that for the detection of air containing 10 ppm, 30 ppm, 60 ppm, 80 ppm, and 99 ppm methane, the errors are 12.75%, 8.18%, 3.37%, 2.46%, and 1.78%, respectively.
- 1
-
Robustness analysis of O-FDE algorithm for dispersion compensation in coherent fiber optical communications
Kun Zhong, Chenping Zeng, Jiaqiang Dong, Baojiang Wang, Dan Li
Abstract:
The overlapping frequency domain equalization (O-FDE) in digital signal processing (DSP) is frequently employed to provide dispersion compensation in long-distance coherent fiber optical communications. However, the change in overlapping symbol length that occurs during the processing of the O-FDE algorithm will typically be influenced by the decision and zero filling of the last subblock, which is harmful to the robustness of the O-FDE algorithm. In this study, with a thorough robustness analysis on changing overlapping symbol length, we present a novel method for decision and zero filling of the last subblock and examine the correspondingly resulting error vector magnitude (EVM) and symbol error ratio (SER) under different values of optical signal-to-noise ratio (OSNR), chromatic dispersion, and overlapped symbol lengths.
- 1
-
Modeling and analysis of pointing error for non-landing vehicle-mounted photoelectric theodolite
Gao Qingjia, Wang Chong, Liu Yanjun, Wang Xiaoming, Wu Tongbang, Zhang Haoyuan
Abstract:
In order to achieve high accuracy measurement for non-landing vehicle-mounted photoelectric theodolite (NVPT) with continuous zoom optical system, a theoretical model of pointing error is presented. Starting with the working mode, the error source of the whole closed loop which affects the pointing error is analyzed. The measurement equation is derived using a coordinate transformation. The pointing error model is then obtained with the help of Mont Carlo method. The error of a NVPT prototype is measured, showing that the pointing error can be predicted accurately. Finally, a high precision NVPT with continuous zoom optical system is successfully designed and analyzed, with an accuracy of 11.9″ at azimuth and 10.9″ at elevation when the optical focal length is set between 1000 mm and 4000 mm.
- 1
-
A complementary binary code based phase unwrapping method
Wenjie Li, Sun Huanghe, Li Fuquan, Wang beibei, Wang Haijian, Gao Xingyu
Abstract:
Phase unwrapping is used to establish the mapping relationship between camera and projector, which is ones of the key technologies in fringe projection profilometry (FPP) based 3D measurement. Although complementary gray code assisted phase unwrapping technology can get a good result on the periodic boundary, it needs more coded images to obtain a high frequency fringe. Aiming at this problem, a complementary binary code assisted phase unwrapping method is proposed in this paper. According to the periodic consistency between the wrapping phase and binary codes, the coded patterns are generated. Then the connected domain strategy is performed to calculate the fringe orders using the positive and negative image binarization. To avoid the mistake near the periodic boundary, complementary binary code inspired by the complementary gray code is proposed. The fringe order correction is also discussed for different situations in the first measured period. Only two binary images are needed in the proposed method, and the fringe frequency is not limited. Both the simulation and experiment have verified the feasibility of proposed method.
- 1
-
A Triple-Band Miniaturized End-Fire Antenna Based on Odd-Mode Spoof Surface Plasmonic Polaritons Waveguide Resonator
Abstract:
A triple-band miniaturized end-fire antenna based on the odd modes of spoof surface plasmonic polariton (SSPP) waveguide resonator is proposed in this paper. To meet the ever increasing demand for more communication channels and less antenna sizes, multi-band antennas are currently under intensive investigation. By a novel feeding method, three odd modes are excited on an SSPP waveguide resonator, which performs as an end-fire antenna operating at three bands, 7.15-7.26GHz, 11.6-12.2GHz and 13.5-13.64GHz. It exhibits reasonably high and stable maximum gains of 5.26, 7.97 and 10.1dBi and maximum efficiencies of 64%, 92% and 98% at the three bands, respectively. Moreover, in the second band the main beam angle shows a frequency dependence with a total scanning angle of 19°. The miniaturized triple-band antenna has a great potential to be applied in wireless communication systems, satellite communication and radar systems.
- 1
-
Research on Conduction Band Offset of CZTSSe solar cell with double absorber layers
hu xaiofeng, xue yuming, dai hongli, wang luoxin, ni meng, bai xin
Abstract:
CZTSSe is considered to be the most potential light-absorbing material to replace CIGS, but the actual photoelectric conversion efficiency of this kind of cell stagnates at 13.6%. One of the reasons for the low efficiency is the high recombination rate of carriers at the interface. In this paper, in order to reduce the carrier recombination, a new solar cell structure with double absorber layers AZO/i-ZnO/CdS/CZTSSe1/CZTSSe2/Mo was proposed, and the optimal conduction band offset of CdS/CZTSSe1 interface and CZTSSe1/CZTSSe2 interface were determined by changing the S ratio of CZTSSe1 and CZTSSe2, and the effect of thickness of CZTSSe1 on the performance of the cell were studied. The efficiency of the optimized single and double absorber layers reached 17.97% and 23.4% respectively. Compared with the single absorber layer structure, the proposed structure with double absorber layers had better cell performance.
- 1
-
Sign Language Quality Screening based on Dual Information Streams
Abstract:
Sign language is the primary way for deaf people to communicate, but hearing people don't understand sign language. Researchers hope to use artificial intelligence technology to help deaf people integrate into society so that hearing people can understand the meaning of sign language. Sign language recognition and translation (SLRT) urgently need large-scale sign language video data to realize its practical application. And the improvement of data quality will also have a positive impact on its recognition and translation effect. To avoid the high cost of manual screening of massive data, this paper proposes a Two Information Streams Transformer (TIST) model to judge whether the quality of a sign language video is qualified. Under the condition of inconsistent sign language style, TIST can use two cross transformers to judge wrong videos where the signer misuses, misses some gestures, wrong gestures order, etc. And TIST also uses the temporal transformer to focus on important frames in the sign language temporal sequence. In addition, this paper also proposes the self-adaptive GCN to enhance the ability to extract sign gestures in skeletal nodes. Experimentally, TIST achieved state-of-the art sign language screening accuracy. To verify that data quality screening is effective in improving sign language recognition, this paper uses VAC as the baseline model. The experimental results show that the screened dataset can achieve better WER than the unscreened dataset.
- 1
-
Multi-wavelength Brillouin Erbium-doped fiber laser with 40GHz frequency shift interval assisted by Sagnac loop filter
Abstract:
A switchable and tunable multi-wavelength Brillouin erbium-doped fiber laser (MWBEFL) is designed and experimentally demonstrated. A Sagnac loop filter is employed as the switcher to obtain the double Brillouin frequency shift (BFS) of 0.172nm (~20GHz) and the quadruple BFS of 0.35nm (~40GHz). The working principles of the proposed laser are theoretically analyzed. The gain in the laser cavity is hybrid gain, which is provided by stimulated Brillouin scattering (SBS) and erbium-doped fiber (EDF). In addition, the 10km single mode fiber (SMF) is not only used as a Brillouin frequency shifter but also provides a certain degree of random distributed feedback. The experimental results show that up to 9 Stokes lines with a wavelength interval of 0.172nm can be obtained. The optical signal-to-noise ratio (OSNR) is greater than 33 dB. The utilization of 980nm pump power and Brillouin pump (BP) power has enhanced the spectral bandwidth of multi-wavelength generation. By adjusting the BP wavelength to investigate the tunability of the fiber laser, the output of 2-5 laser channels can be realized corresponding to 20 nm wavelength range. This approach is simple and can be employed for the microwave generation of other frequency ranges subject to the filtering shift of the Sagnac loop.
- 1
-
Point-Voxel Dual Transformer for LiDAR 3D Object Detection
Abstract:
Recently, the research on transformers in deep learning makes a tremendous progress in nature language processing and computer vision. Due to the inherent invariant permutation, transformers provide solutions to unordered points problems faced by deep learning in point cloud object detection. In this paper, a two-stage LiDAR 3D object de-tection framework is presented, namely Point-Voxel Dual Transformer (PV-DT3D), which is a transformer-based method. In the proposed PV-DT3D, point-voxel fusion features are used for proposal refinement. Specifically, in the PV-DT3D, keypoints are sampled from entire point cloud scene and used to encode representative scene features via a proposal-aware voxel set abstraction module. Then according to the generated proposals by region proposal networks (RPN), the internal encoded keypoints are fed into dual transformer encoder-decoder architecture. In 3D object detection, for the first time, the proposed PV-DT3D takes advantages of both pointwise transformer and channel-wise architecture for capturing contexual information from the perspective of spatial and channel dimen-sions. Experiments conducted on the highly competitive KITTI 3D car detection leaderboard show that, the PV-DT3D achieves superior detection accuracy among state-of-the-art point-voxel-based methods.
- 1
-
Wavelength injection locking actively Q-switched random fiber laser based on random phase-shifted fiber Bragg grating and EOM
ZHANG Bo, PAN Honggang, SONG Dianyou, CHEN Zhi pan, CHEN Chunqi, LI Rupeng
Abstract:
In this paper, an actively Q-switched wavelength injection-locking random fiber laser based on random phase-shifted fiber Bragg grating is proposed, and the performance of the laser is verified by experiments. Within the reflection bandwidth of random phase-shifted fiber gratings from 1549.2 to 1549.9nm, different laser modes with stable central wavelength and peak power can be selected by changing the wavelength of the injected light. The power fluctuation within one hour is less than 0.1dBm, and the central wavelength drift is less than 0.02nm. When the pump power increases from 90mW to 300mW, the pulse width decreases from 3.2μs to 1.5μs, and the pulse repetition frequency is 20kHz. The Random fiber laser (RFL) can reach a stable locking state at the lowest pump power of 100mW and the lowest injection power of 3dBm. When the wavelength is locked, the output pulse is a single pulse. On the contrary, the unlocked output pulse is multi-pulse. The laser has the characteristics of high wavelength tunability in the reflection range of random phase shift fiber bragg grating (RPS-FBG), and it can be an ideal light source in the fields of laser imaging and pulse coding.
- 1
-
A Nano-Plasmonic HMIM Waveguide Based Concurrent Dual-Band BPF Using Circular Ring Resonator
Miriyala Sridhar, Surendra Kumar Bitra, T S N Murthy, KOPPIREDDI PADMARAJU4
Abstract:
This article analyzes the transmission line characteristics of plasmonic Hybrid Metal Insulator Metal (HMIM) waveguide, circular ring resonator (CRR) based dual-band band-pass filters with two transmission poles in both pass-bands in the optical regime using coupled line feed. The transmission line characteristics of an HMIM waveguide, such as characteris-tic impedance (ZPV), effective refractive index (Neff) and propagation length (Lspp) have been obtained by using full wave simulation. Using basic HMIM slot waveguide, a CRR with periodic loading of double and triple ring CRR are numerically analyzed. Two input ports have been used for excitation which are located at the separation of 180? positions along the CRR, and is coupled with the ring by parallel coupled lines, producing the dual pass-bands with the synchro-nous excitation of two transmission poles. The proposed Double Ring Dual-Band Band Pass Filter (DR-DB-BPF) offers 35 dB extinction ratio (ER), 299.69 nm free spectral range (FSR) and narrow band Full Width Half Maximum (FWHM) of 78.057-112.43 nm. The Triple Ring DB-BPF (TR-DB-BPF) has 22.5 dB ER, FSR of 292.18 nm and FWHM is 42.751-59.58 nm. The proposed filters are very useful in the development of dual-band filters for electronic photonic integrated circuits (EPICs), as the optical signals are filtered at two wavelengths simultaneously.
- 1
-
Effect of substrate pretreatment on HMDSN plasma deposited thin films properties
Saker Saloum, S. A. Shaker, R. Hussin, M. N. Alkafri, A. Obaid, M. Alsabagh
Abstract:
Silicon substrates were plasma treated before being coated by plasma-polymerized Hexamethyldisilazane (pp-HMDSN) thin films. The pretreatments included oxygen, SF6 and argon plasmas. The effect of these pretreatments on the properties of deposited thin films was studied, including film thickness, morphology and photoluminescence (PL). The deposition of pp-HMDSN thin films was performed using plasma enhanced chemical vapor deposition (PECVD). It was found that the substrate pretreatment induces an increase of film thickness, and the morphology of the deposited thin film follows that of the treated substrate, while the intensity of PL increases due to change of nanostructure accompanied with roughness and thin film thickness increase.
- 1
-
The lateral photovoltaic effect in the Ni-SiO2-Si structure with bias
Abstract:
We had designed a clamping device to study lateral photovoltaic effect (LPE) in Ni-SiO2-Si structure with bias due to the appropriate barrier height, the LPE have a prominent sensitivity and linearity with 532 nm wavelength laser. The transient response time is 450 μs and the relaxation time is 2250 μs in Ni-SiO2-Si structure without bias. The LPE sensitivity have a significant improvement with bias. The transient response time is 6 μs and the relaxion time is 47 μs with -7 V bias. Not only improve the LPE sensitivity, but also increase response speed with bias. The research shows that the Schottky barrier structure can improve the sensitivity and linearity of LPE with bias effectively and thus it can be used in position sensitive sensors.
- 1
-
Recent advances in MXene for terahertz applications
Abstract:
Since first synthesized in 2011, MXene has attracted extensive attention in many scientific fields as a new two-dimensional material because of its unique physical and chemical properties. Over the past decade, in particular, MXene has obtained numerous exciting achievements in the field of terahertz applications. In this review, we first briefly introduce the MXene materils, such as the basic structure and main fabrication processes of MXene. Then, we summarize the recent applications of MXene materials in various terahertz research areas, including terahertz modulation, terahertz absorption, terahertz shielding, terahertz communication, terahertz detection and terahertz generation, in which the representative results are presented. Finally, we give an outlook on the future research directions of MXene materials and their potential applications.
- 1
-
Improved remote sensing image target detection based on yolov7
Abstract:
Remote sensing images are taken at high altitude from above, with complex spatial scenes of images and a large number of target types. The detection of image targets on large scale remote sensing images suffers from the problem of small target size and target density.The paper here proposes an improved model for remote sensing image detection based on YOLOv7. First, the small-scale detection layer is added to reacquire tracking frames to improve the network's recognition ability of small-scale targets; then Bottleneck Transformers are fused in the backbone to make full use of the CNN+Transformer architecture to enhance the feature extraction ability of the network; after that, the CBAM attention mechanism is added in the head to improve the model's ability of small-scale target; Finally, the non-maximum suppressed NMS of YOLOv7 algorithm is changed to DIOU_NMS to improve the detection ability of overlapping targets in the network. The results show that the method in this paper can improve the detection rate of small-scale targets in remote sensing images and effectively solve the problem of high overlap, and is tested on the NWPU-VHR10 and DOTA datasets, and the accuracy of the improved model is improved by 6.3% and 4.2%, respectively, compared with the standard yolov7 algorithm.
- 1
-
Coherence analysis of noise-like pulses generated by an erbium-doped fiber mode-locked laser
ZIYI ZHANG, CHUNCAN WANG, JING LI, PENG WANG
Abstract:
The noise-like pulses (NLPs) with tunable fraction of the pedestal height in the whole intensity autocorrelation (AC) trace are numerically demonstrated in the designed erbium-doped fiber (EDF) mode-locked laser, which contains the saturable absorber (SA) with nonlinear polarization rotation (NPR), sinusoidal-shaped or Gaussian-shaped filter, two segments of EDFs, and two pieces of single-mode fibers with normal dispersion and anomalous dispersion, respectively. The pedestal ratio of the intensity AC trace can be tuned by changing the gain saturation energies of EDFs. The results show that when the net cavity dispersion is 1.06 ps2, the tuning range of the pedestal ratio for the NLPs can reach its maximum values, which are 0.51-0.89 and 0.58-0.88 for the sinusoidal-shaped and Gaussian-shaped filters, respec-tively. In addition, an appropriate choice of filter bandwidth is also conducive to obtain a wide range of the tuning pedestal ratio for the intensity AC trace.
- 1
-
A Simplified Dispersion-compensation Microstructure Fiber with Seven cores
Li Wenchao, Wang Chao, Yang Hongda, Hang Ying, Wang Wei
Abstract:
In order to compensate the dispersion accumulated in a single mode fiber for higher communication capacity, a simplified dispersion-compensation microstructure fiber with seven cores is proposed in this paper. The fiber’s cladding is made of pure silica without air holes, and its outer cores are composed of 6 germanium up-doped cylinders, which has the ad-vantage of simple structure. The Finite Element Method and Beam Propagation Method is used to study the properties of the fiber, and the relationship between the structure parameters of the fiber and the dispersion, as well as the phase matching wavelength, is obtained. By optimizing the structural parameters of the fiber, the dispersion of the fiber can reach -5291.47ps/(nm-km) at 1550nm, and the coupling loss to the conventional single-mode fiber is only 0.137dB. Compared with the conventional dispersion-compensation fiber, the fiber has lots of advantages, such as single mode transmission, easy to fabricate and low coupling loss with traditional single mode fiber, etc.
- 1
-
Discriminative low-Rank embedding with manifold constraint for image feature extraction and classification
Abstract:
Robustness to noise, outliers, and corruption is an important issue in image feature extraction. A discriminative low-rank embedded image feature extraction algorithm is proposed in this paper to address this problem. Firstly, manifold con-straints are introduced based on the low-rank embedding (LRE) approach to capture the geometric structure of the local manifold, taking into account both local and global information. Secondly, a discriminant analysis term is also introduced to obtain global discriminant information and learn the optimal projection matrix for data dimensionality reduction. Finally, the test samples are projected into a low-dimensional space for classification. Numerical experiments show that the classi-fication accuracy of the method proposed in this paper is 95.62%, 95.22%, 86.38%, and 86.54% on the ORL, CMUPIE, AR, and COIL20 datasets, respectively, which is better than that of the dimensionality reduction-based image feature ex-traction algorithm.
- 1
-
Steel surface defect detection based on lightweight YOLOv7
shi tao, wu rong xin, zhu wen xu, ma qing liang
Abstract:
Aiming at the problems of low detection efficiency and difficult positioning of traditional steel surface defect detection methods, a lightweight steel surface defect detection model based on YOLOv7 is proposed. First, a CSS Block module is proposed, which uses more lightweight operations to obtain redundant information in the feature map, reduces the amount of computation, and effectively improves the detection speed. Secondly, the improved SPPCSPC structure is adopted to ensure that the model can also pay attention to the defect location information while predicting the defect category information, obtain richer defect features. In addition, the convolution operation in the original model is simplified, which significantly reduces the size of the model and helps to improve the detection speed. Finally, using EIOU Loss to focus on high-quality Anchors, speed up convergence and improve positioning accuracy. Experiments were carried out on the NEU-DET steel surface defect data set. Compared with the original YOLOv7 model, the number of parameters of this model was reduced by 40%, the FPS reached 112, and the average accuracy reached 79.1%., the detection accuracy and speed have been improved, which can meet the needs of steel surface defect detection.□
- 1
-
A Grapefruit Microstructure Fiber Temperature Sensor Coated with Liquid Crystal based on Waist-enlarged Taper
WANG Feng, LI Jiaxuan, ZHANG Rongjing, FU Xinghu
Abstract:
In this paper, a Grapefruit microstructure fiber(GMF) temperature sensor coated with liquid crystal(LC) based on waist-enlarged taper is proposed and fabricated, and its temperature sensing characteristics are analyzed. The waist-enlarged taper is formed at the fusion point between single mode fiber(SMF) and GMF. The capillary glass tube is sleeved outside GMF, LC is filled into the capillary glass tube, and its two ends are finally sealed to form a sensor. The experimental results show that when the length of GMF is 2.5 cm, the temperature sensitivity of the sensor can reach up to 195.3 pm/℃ in the range of 30~90 ℃, and it has a good stability for reuse. Thereby, it can be used in biochemical, industrial production and other temperature detection areas.
- 1
-
An electrically controlled tunable photonic crystal filter based on thin-film lithium niobate
Yifan Wang, Yuan Yao, Hao Zhang, Bo Liu, Shaoxiang Duan, Haifeng Liu, Wei Lin
Abstract:
In this paper, we present an electrically controlled tunable photonic crystal filter based on thin-film lithium niobate. The filter incorporates a photonic crystal microcavity structure within the straight waveguide, enabling electronic tuning of the transmitted wavelength through added electrode structures. The optimized microcavity filter design achieves a balance between high transmission rate and quality factor, with a transmission center wavelength of 1551.6nm, peak transmission rate of 96.1%, and quality factor of 5054. Moreover, the filter can shift the central wavelength of the transmission spectrum by applying voltage to the electrodes, with a tuning sensitivity of 13.8pm/V. The proposed tunable filter adopts a simple-to-fabricate air-hole structure and boasts a compact size (length: 11.57μm, width: 5.27μm, area: 60.97μm2), making it highly suitable for large-scale integration. These features make the filter promising for broad applications in the fields of photonic integration and optical communication.
- 1
-
Numerical calculation of z-scan measurements for nonlinear media with large phase shift
Mohammad Dergam Zidan, A. Allahham
Abstract:
We have reported the characteristics of a z-scan for the Poly(azaneylylidene-acylene) "DAZA" polymer as nonlinear medium with a large nonlinear phase shift using CW laser beam. It has been verified that the Fresnel diffraction model is applicable for analyses of z-scan measurements with DAZA polymer at high power. It was found that z-scan curves with peak-to-valley features appear as the applied light intensity increases in the case of a large nonlinear phase shift. The z-scan experiments were carried out of DAZA polymer by a CW laser to verify the theoretical calculations in the case of a large nonlinear phase shift model. Our results show good agreements between the experimental data and the proposed theoretical model.
- 1
-
AWG-based large dynamic range fiber Bragg grating interrogation system
Li Shufeng, Yuan Pei, Li Ke, Li Ting, Zhu Lianqing
Abstract:
Arrayed waveguide gratings (AWG) are extensively employed in fiber Bragg grating (FBG) interrogation systems due to their compact size, lightweight nature, and excellent interrogation performance. The resolution and total measurement range of AWG-based FBG interrogation systems are constrained by the output properties of AWG. We proposed an AWG-based large dynamic range interrogation system. The temperature dependence of AWG is exploited to achieve continuous interrogation. The test results show that the interrogation system has a dynamic range of 28.67 nm, an inter-rogation accuracy better than 25 pm, and a wavelength resolution of 6 pm.
- 1
-
Research on the identification of the production origin of Angelica dahurica using LIBS technology combined with machine learning algorithms
Sun Jiaxing, Li Honglian, YAO Yuhang, YAN Qiongyan, Li Xiaoting
Abstract:
Different production origins of Angelica dahurica have varying pharmacological effects. In order to achieve fast and accurate identification of the production origin of Angelica dahurica, this study combines laser induced breakdown spectroscopy (LIBS) technology with machine learning algorithms to identify the original production origin of Angelica dahurica. Sliced samples of Angelica dahurica were taken from four regions: Hebei, Henan, Zhejiang, and Sichuan. The spectral data from the sliced samples were used as features, and different algorithms including support vector machine (SVM), backpropagation (BP) neural network, genetic algorithm-backpropagation (GA-BP) neural network, particle swarm optimization-backpropagation (PSO-BP) neural network, convolutional neural network (CNN), and CNN-SVM were employed to classify the origin of Angelica dahurica samples. The results show that the average prediction accuracy of the BP, GA-BP, and PSO-BP algorithms reached 89.64%, 89.66%, and 89.93% respectively. The average prediction accuracy of the SVM, CNN, and CNN-SVM algorithms reached 89.92%, 90.32%, and 90.53% respectively. The average prediction accuracy improved when the two algorithms were combined, and the CNN-SVM algorithm showed a 44% increase in the lowest prediction accuracy compared to the SVM algorithm. Overall, the combination of the CNN-SVM algorithm and LIBS technology demonstrated the best performance for identifying the origin of Angelica dahurica, a traditional Chinese medicinal herb, and can provide reference for the origin identification of medicinal materials.
- 1
-
Research on a Fiber Sensing System for Metal Ion De-tection Based on SPF-PCF-SPF Structure and Coated LPFG
Abstract:
In order to provide a method for accurately detecting the concentration and types of heavy metal ions in water, a fluid ion detection system is designed. It consists of a side-polished fiber-assisted fluid structure and a long-period fiber grating coated with a metal chelating agent membrane. In this study, both theoretical and experimental inves-tigations are conducted to examine the sensing characteristics of the system towards copper ion and iron ion solu-tions. The results demonstrate that, under the premise of ensuring solution flow, the system can achieve specific identification of different types of heavy metal ions. Furthermore, it exhibits concentration sensing sensitivities of 9.23?104 mL?nm/mol and 7.13?104 mL?nm/mol for copper sulfate and ferric chloride solutions, respectively. Therefore, this sensing system offers the potential for real-time detection of metal ions.
- 1
-
An Improved Construction Algorithm of Polar Codes Based on the Frozen Bits
Abstract:
In order to improve the problems that the Minimum Hamming Weight (MHW) of the polar codes of the traditional Gaussian Approximation (GA) construction is small and its performance is not good enough, an improved channel construction algorithm of polar codes based on frozen bits is proposed by combining the construction of the RM code to effectively increase the MHW and analyzing the correcting and checking functions of the frozen bits in the Successive Cancellation List (SCL) decoding. The construction algorithm selects the channel with the smaller row weight corresponding to the information channel in the channel construction stage, and some channels are set as the frozen channels under the proposed frozen channel setting principle. So the proposed construction algorithm not only eliminates the channels with the smaller row weight and optimizes the distance spectrum of polar codes, but also makes full use of the checking ability of the frozen bit in SCL decoding to improve the error correction performance of polar codes. The polar codes constructed by this algorithm is named as FRM-Polar codes. The simulation results show that the proposed FRM-Polar codes have a larger performance gain than the RM-Polar codes and the polar codes constructed by GA under the different code-length. In addition, the proposed construction algorithm has the same complexity as the construction algorithm of the RM-Polar codes.
- 1
-
Photonic sensor with radio frequency power detection for body pressure monitoring
Abstract:
A photonic sensor with radio frequency (RF) power detection for body pressure monitoring is proposed. The sensor based on two FBGs can transfer the wavelength shift to the change of RF power. The pressure can be measured by modulating and processing one single frequency RF signal. The theoretical analysis and experimental results of the photonic sensor are presented and discussed. The pressure sensitivities are acquired with 2.62e-5mW/kPa at 2.14GHz, 2.46e0-5mW/kPa at 2.21GHz, 2.81e-5 mW/kPa at 2.37GHz and 3.02e-5 mW/kPa at 2.45GHz, respectively. Furthermore, the pressure measurements of pressed body parts are also obtained by the sensor.
- 1
-
Antibody-Modified Integrated Microfluidic Terahertz Biosensor for Detection of Breast Cancer Biomarkers
Abstract:
Breast cancer is the most common malignant tumor in women, which seriously threatens the physical and mental health of women worldwide. The existing detection methods have problems such as large sample consumption, time-consuming sample preparation, expensive equipment, and low sensitivity. In order to solve these problems, this paper proposes a method for quickly detecting breast cancer using surface-functionalized terahertz metamaterial biosensors. The use of PIK3CA-modified sensors enhances the detection sensitivity and specificity of exosomes. Based on the red shift of the sensor absorption peak caused by exosomes, breast cancer patients can be distinguished from healthy controls. This study demonstrates that exosome detection is effective for the repeatable and non-invasive diagnosis of breast cancer patients. The terahertz metamaterial biosensor designed in this paper has high specificity, repeatability, and sensitivity, and has great potential for application in the development of modern diagnostic instruments.
- 1
-
Q-Switched Giant Pulsed Erbium-doped All-fibre Laser with V2ZnC MAX Phase Saturable Absorber
M Musthafa, Azura Hamzah, Wei Ling, Rosol, Mohamed, Harun
Abstract:
MXenes, drawn from MAX phases are special two-dimensional substances with numerous advantages in nonlinear optics, specifically in giant and ultrashort pulsed-laser applications. Ti3C2Tx, and Ti2CTx nanosheets however, rapidly deteriorate under ambient conditions, limiting their applications. This paper demonstrates how excellent modulation depth of one of the MAX phase compounds V2ZnC makes it a brilliant saturable absorber in passively Q-switched all-fibre pulsed lasers—integrated such that a 16.73-µm V2ZnC-PVA thin film acts as saturable absorber in the laser. Saturable and non-saturable absorptions were found to be 13.2 % and 10.47 %, while saturation optical intensity and modulation depth were 6.25 kW/cm2 and 12.43 %, respectively, illustrating the optical nonlinearity. The superiority of MAX-PVA, fabricated in four distinct ratios was demonstrated by the fact that it self-starts a giant pulsed laser at pump power as low as 22.5 mW and firmly accomplished 120.6 kHz repetition rate with a pulse width of 2.08 µs. It is a fine SA for the use of pulsed-laser production using all-fibre laser due to fabrication simplicity and great optical, thermophysical, and mechanical qualities.
- 1
-
Research of Surface-Enhanced Raman Scattering on Ag@PMBA@C@Au hybrid nanoparticles
Abstract:
Monodispersed, biocompatible, and readily-functionalized hybrid reporter-embedded core-shell nanopartilces have been prepared in a simple route. This composite offers a potential platform for immunochemical detection using surface-enhanced Raman scattering (SERS) due to their high sensitivity, good stability and biocompatiblity. This also provides a new platform for insight into SERS enhancement mechanism.
- 1
-
Numerical investigation of supercontinuum generation and optical frequency combs in a SiN-based PCF with high nonlinear coefficient
Saeed Olyaee, Alizadeh, Seifouri
Abstract:
In this paper, a photonic crystal fiber (PCF) with a dispersion-engineered and high nonlinear coefficient has been designed for supercontinuum generation and frequency comb generation. The proposed PCF has a Si3N4 rod in the core. This rod increases the contrast between the core and the cladding which provides more optical confinement in the core. This confinement reduces the effective mode area of PCF and thus increases the nonlinear coefficient. The effective mod area and the nonlinear coefficient are obtained "0.8" 14〖 μm〗^2 and 25 W-1m-1, respectively. By varying different parameters for dispersion engineering, a suitable dispersion profile for the structure has been obtained so that the proposed PCF has two zero dispersion wavelengths (ZDWs) at 900 nm and 1590 nm. By injecting an input pulse at a wavelength of 1555 nm and a power of 1 kW with a duration of 50 fs to the designed PCF with a length of 4 mm, the output spectrum is broadened in the range of 800 nm to 3500 nm. For frequency comb generation by the four-wave mixing method, phase matching conditions must be provided, and for that, the pumped wavelength must be in the anomalous dispersion regime and near to zero dispersion wavelength. As a result, two continuous wave laser pumping at the wavelengths of 1551 nm and 1558 nm have been injected into the PCF and optical frequency combs with a pulse width of 1 nm and a free spectral range of 7 nm has been obtained.
- 1
-
E-MobileNeXt：Face expression recognition Model based on improved MobileNeXt
Abstract:
In response to the high complexity and low accuracy of current facial expression recognition networks, this paper proposes an E-MobileNeXt network for facial expression recognition. E-MobileNeXt is built based on our proposed E-SandGlass block. In addition, we also improve the overall performance of the network through RepConv and SGE attention mechanisms. The experimental results show that the network model improves the expression recognition accuracy by 6.5% and 7.15% in RAF-DB and CK datasets, respectively, while the parameter and floating-point operations decreased 0.79M and 4.2M compared with MobileNeXt
- 1
-
Research on denoising of joint detection signal of water quality with multi-parameter based on IEEMD
Abstract:
An improved ensemble empirical mode decomposition (IEEMD) is suggested to process water quality spectral signals in order to address the issue that noise interference makes it difficult to extract and evaluate water quality spectral signals. This algorithm effectively solves the problems of modal mixing, poor reconstruction accuracy in the empirical mode decomposition (EMD), and a large amount of calculation in the ensemble empirical mode decomposition (EEMD). Based on EEMD, IEEMD firstly preprocesses the original water quality spectral signals, then performs savitzky-golay (S-G) smoothing on the decomposed effective intrinsic mode function (IMF) components, and finally reconstructs them to obtain the denoised signals. Water sample data at different concentrations can be accurately analyzed based on the noise-reduced spectral signals. In this paper, three water quality parameters are used as research objects: benzene (C6H6), benzo(b)fluoranthene (C20H12), and chemical oxygen demand (COD). The original water quality multi-parameter (C6H6, C20H12, COD) spectral signals were subjected to denoising based on the IEEMD and the water quality multi-parameter joint detection technology. The signal-to-noise ratio (SNR) and the correlation coefficient (R2) of the fitted curves obtained from the processing of the IEEMD were compared and analyzed with those obtained from the processing of the EMD and the EEMD. The experimental results show that the SNR of the spectral signals and the R2 of the fitting curve in three water quality parameters have been significantly improved. Therefore, the IEEMD effectively improves the phenomenon of modal mixing, reduces the amount of calculation, improves the reconstruction accuracy, and provides an important guarantee for the effective extraction of multi-parameter spectral signals of water quality.
- 1
-
FORMALDEHYDE CONCENTRATION MEASUREMENT BASED ON OPTICAL BUNDLE FIBER AND UV-VIS LIGHT SOURCE*
M. Yasin, M. F. F. PRADIPTA, H. TRILAKSANA, M. ZULKARNAEN, S. W. Harun, S. K. LIAW
Abstract:
We have successfully demonstrated a new sensor based on an optical fiber bundle for measuring formaldehyde concen-tration. Our sensor was tested using two different methods based on displacement tuning and reflection measurement in conjunction with a UV light source. The sample was illuminated with the UV-visible laser (OSL 2 Thorlabs), and the power output of the backscattered radiation was found to decrease linearly with respect to the formaldehyde concentration. Our sensor exhibits excellent sensitivity, linearity, and stability as it was tested for formaldehyde concentrations ranging from 0% to 5%. The displacement tuning method exhibited the sensitivity of 0.031 μW/% with 91% linearity while the reflection scheme provided a a sensitivity of 2634.3 counts/% with 98% linearity. In addition, our sensor is non-contact, which means that the sample and probe are protected from damage and harm. These results demonstrate the potential of our fiber bundle sensor as a reliable tool for monitoring formaldehyde concentrations in food safety applications in the future.
- 1
-
Observation of regular pulse train in a narrowband optoelectronic oscillator
Zhao Chunbo, Tuo Zhuoyue, Yao Jiali, He Yuling, Zhai Shenghua, Meng Yansong
Abstract:
We have experimentally observed a new operating state of a regular pulse train in a narrow-band optoelectronic oscillator system, where the DC bias of the Mach-Zehnder modulator is set at the maximum value of the transmission transfer function instead of the usual quadrature point. The observed quasi-steady-state pulse train is distinctly periodic, with a period of 10.5 ?s and a center frequency of 10 GHz, and resembles a mode-locked optoelectronic oscillator in its waveform. The formation of regular pulses here may arise from the dynamic balance of non-linearity and narrowband filter effects, with the transient characteristics of the pulses arising mainly from instabilities between the gain and cavity loss. Our results are of great importance for deepening the understanding of the non-linear dynamical processes in OEO systems.
- 1
-
A Multi-scale Detection Approach Based on Deep Neural Networks for Multi-scale Object Detection
Zhu Yongchang, Tong Jigang, Yang Sen, Wang Zenghui
Abstract:
The effectiveness of deep learning networks in detecting small objects is limited, thereby posing challenges in addressing practical object detection tasks. In this research, we propose a small object detection model that operates at multiple scales. The model incorporates a multi-level bidirectional pyramid structure, which integrates deep and shallow networks to simultaneously preserve intricate local details and augment global features. Moreover, a dedicated multi-scale detection head is integrated into the model, specifically designed to capture crucial information pertaining to small objects. Through comprehensive experimentation, we have achieved promising results, wherein our proposed model exhibits a mean average precision (mAP) that surpasses that of the well-established YOLOv7 model by 1.1%. These findings validate the improved performance of our model in both conventional and small object detection scenarios.
- 1
-
InGaN multi quantum well based light-emitting diodes with indium composition gradient InGaN quantum barriers
Abstract:
To improve the internal quantum efficiency (IQE) and light output power of InGaN light-emitting diodes (LED), we proposed an In-composition gradient increase and decrease InGaN quantum barrier structure. Through analysis of its P-I graph, carrier concentration, and energy band diagram, the results showed that when the current was 100 mA, the In-composition gradient decrease quantum barrier (QB) structure could effectively suppress electron leakage while improving hole injection efficiency, resulting in an increase in carrier concentration in the active region and an improvement in the effective recombination rate in the quantum well (QW). As a result, the IQE and output power of the LED were effectively improved.
- 1
-
Dynamic behavior analysis, color image encryption and circuit implementation of a novel complex memristive system
Li Xiong, Wang Xuan, Zhang Xinguo, He Tongdi
Abstract:
This paper is devoted to introduce a novel four-dimensional memristor-involved system and its applications in image encryption and chaotic circuit. The typical dynamical behaviors of the memristor-involved system are explored, such as chaotic phase potraits, Lyapunov exponent spectrum (LES), bifurcation diagram (BD) and complexity analysis. Then a color image encryption algorithm is designed. In this algorithm, the sequences generated by the four-dimensional memristor-involved system are used in scrambling and diffusion algorithm for three channels. The algorithm analysis results based on key space, key sensitivity, information entropy, histogram distribution, correlation coefficients, data loss and noise attacks indicate that the proposed algorithm can improve the security of the color image encryption algorithm. Finally, the memristor-involved chaotic circuit is implemented by using some discrete components. The experimental results of hardware circuit are consistent with the Multisim simulation results and the numerical simulation results. The research results have certain universality and portability, and can provide technical support for the subsequent analysis of other nonlinear circuits and the application of chaotic secure communication.
- 1
-
Research on wind turbine blade damage based on pre-stressed FBG strain sensors
Abstract:
Wind turbine blades are one of the core components that capture wind energy in a wind power generator. As the service time of wind turbines increases, the safety of the blades gradually de-creases. Therefore, a chip-type pre-stressed FBG strain sensor was designed for real-time moni-toring. Combined with simulation analysis, the structure of the sensor was optimized. Through calibration experiments, it was found that the pre-stressing process increased the measurement range of the sensor, ensured a consistent overall linearity, and avoided the possible hysteresis phenomenon during compression. The final sensitivity of the sensor was determined to be 1.970pm/με, with a linear fitting coefficient of 0.999. Finally, the sensor was used to monitor the wind turbine blades and it was found that the strain change curve of the root of a normally functioning blade is a sine curve, which provides a certain reference value for judging whether the blade is damaged in the future.
- 1
-
Seawater temperature and salinity sensing based on in-fiber Michelson-Fabry-Perrot hybrid interferometer employing frequency domain decomposition method
WANG Jia-hui, Wei Xian, ZHU Yan, SHEN Yue, PANG Lipeng, DUAN Shao-xiang
Abstract:
In this letter, we propose an in-fiber Michelson-Fabry-Perrot (M-FP) hybrid interferometer for the simultaneous meas-urement of seawater temperature and salinity. The sensor head consists of two parallel hetero Fabry-Perrot cavities fabri-cated on the end face of the twin core fiber. A fiber fusion taper is used to split and recouple the light in the two cores. In this case, the Vernier effect can be obtained which can greatly enhance the sensitivity and solve the problem of tempera-ture cross-sensitivity. Different from the traditional demodulation method based on envelop detection, we employed fre-quency domain decomposition method (FDDM) to demodulate the sensing signal. The simulation results indicate that the proposed sensor has high sensitivity to salinity and temperature. Thanks to the merits of high sensitivity, ease of fabrica-tion and small footprint, the proposed seawater temperature and salinity sensor would have potential applications in ma-rine science, food industry and ocean ranching.
- 1
-
The electrochromism and photoeletrochemical performance of WO3/Au composite film electrodes
chunlei liu, jikai yang, haorui liu, yiming zhao
Abstract:
WO3/Au composite film electrode was prepared by hydrothermal combined electrodeposition method. The samples were characterized by SEM, EDS and XRD and the results showed that WO3/Au composite film was synthesized. Electrochemical and spectral measurements were carried out to obtain the electrochromic conversion time, reversibility, coloration efficiency and transmittance of the samples. The photocurrent and photoelectric catalysis degradation efficiency were carried out to obtain photocurrent and photoelectric catalysis of the samples. The results show that compared with pure WO3 nanoblocks, WO3/Au composite film improves the electrochromic property, photocurrent and photoelectric catalysis activity. Among them, WO3/Au composite film prepared by depositing Au nanoparticles in 80 s showed the highest electrochromic property, photocurrent and photoelectric catalysis activity. Meanwhile, the photoelectric catalysis activity of the composite film is higher than its direct photocatalysis and electrocatalysis activity.
- 1
-
An All-Optical 1*2 De-Multiplexer based on Two-Dimensional Nonlinear Photonic Crystal Ring Resonators
Esmat Rafiee, Fatemeh Abolghasemi
Abstract:
In this work, a new configuration of an all-optical nonlinear de-multiplexer gate based on two-dimensional photonic crystals is proposed. The gate is considered in the double-ring resonator shaped structure of Silicon rods. In order to have a more functional structure, some defect rods made of nonlinear materials were positioned in the structure. For consider-ing the functionality of the structure, photonic band gap (PBG), field distribution and transmitted power spectra are in-vestigated. Plane wave expansion and finite-difference-time-domain methods are utilized for extracting the PBG and field distribution diagrams. The remarkable dimension, bit rate, maximum intensity and contrast ratio of 116.64μm2, 3.125Tbit/s, 97% and 40.2dB are obtained, respectively which make the gate an appropriate candidate for utilization in optical integrated circuits.
- 1
-
Detection of Pb in Tieguanyin tea and ash by laser-induced breakdown spectroscopy
Li Jingwen, Yao Lixing, Shen Li, Wang Cong
Abstract:
In this work, laser-induced breakdown spectroscopy (LIBS) was applied for the detection of Pb in Tieguanyin tea and ash. Firstly, the Tieguanyin tea and ash containing Pb were prepared, and the difference of intensities of Pb I spectral lines before and after the ashing treatment was studied. It was found that the intensities of Pb I lines increased by 30 times and the standard deviation of background signal decreased by 41% after the ashing treatment. Therefore, the enrichment of Pb element by ashing treatment was used to detect Pb in tea with high sensitivity. Then, the cali-bration curve of Pb was established using spectral lines without self-absorption, and the determination coefficients (R2) for the linear fitting of calibration curve was 0.9799. Finally, it was found that the limit of detection of Pb was 233.8 ppb. Compared with the results of direct detection of Pb in tea of other works, the enrichment of Pb by ashing treatment improved the detection sensitivity of Pb by about 200 times. In addition, this method can be applied to the high sensitivity detection of other heavy metals, such as Cr, Cd, Hg, etc. in plants, Chinese herbal medicine, flour, rice, coal and other solid materials.
- 1
-
DUAL WAVELENGTH ALL-FIBER Q-SWITCHED FIBER LASER USING A BISMUTH DOPED FIBER AS THE SATURABLE ABSORBER
Sulaiman Wadi Harun, ARUNA GHOSH, ABDULKADIR MUKHTAR DIBLAWE, UTTAM KUMAR SAMANTA, SHYAMAL DAS, ANIRBAN DHAR, AHMAD HAZIQ AIMAN ROSOL, MOH YASIN, Mukul Chandra Paul
Abstract:
In this letter, a passively Q-switched all-fiber laser is demonstrated using a 10 cm long Bismuth-doped fiber as a saturable absorber (SA). The dual wavelength operation was obtained due to the nonlinear effect inside the fabricated BDF, which has a high germanium content. Stable Q-switched pulses were obtained at the dual synchronous wavelengths of 1530.1 nm and 1531.1 nm. When the pump power is tuned from 105.3 to 191.0 mW, the repetition rate can be varied from 82.6 to 117.6 kHz. The maximum pulse energy and average output power were 83.4 nJ and 9.8 mW, respectively while the minimum pulse width was 8.5 μs at the maximum pump power of 191.0 mW. To the best of our knowledge, this is the first report on a dual-wavelength Q-switched laser based on a BDF SA. Our results indicate that BDF could be a promising alternative optical modulator for pulsed fiber laser application.
- 1
-
Research and Analysis of Brillouin distributed Sensing System based on Quasi-single-mode few-mode Fiber
Li Yongqian, Liu Zijuan, Zhang Lixin, Tian Min, Fan Haijun
Abstract:
A distributed fiber sensor was created by splicing two SMFs using the FMF technique. A BOTDA system was developed to measure the sensor's temperature and bending performance. Two-mode and four-mode step FMFs were combined to splice the few-mode segment. The results indicate that the temperature response coefficients of the few-mode segment are only slightly higher than those of the connected single-mode segment, measuring at 1.13MHz/℃ and 1.12 MHz/℃, respectively. The minimum bending radius for the sensor is 0.9cm, and the four-mode bending response curve is superior to that of the two-mode, proving that 4-SI-FMF offers better-bending sensitivity.
- 1
-
Analysis of light propagation characteristic in the aero-optic flow field of cone-headed vehicle with side window
Xu Liang, Zhou Liye, Wang Luyang, Zhao Shiwei, Wang Tao
Abstract:
In this article, aero-optic imaging deviation research is carried out for infrared-guided vehicle with cone-head side window, with a focus on the propagation characteristics of light in an aero-optic flow field. When the light entering the aero-optic flow field from the free-stream should be close to the normal, but numerous data indicate that the light is refracted away from the normal. This paper divides the aero-optic flow field into two parts and uses the gas density distribution in the aero-optic flow field to propose the hypothesis that there are two modes of refraction when light propagates through the flow field. The result show that light propagates from the optically denser medium to the optically thinner medium after passing through the shock wave and eventually leads to refraction away from the normal when the light enters the aero-optic flow field.
- 1
-
Deep Learning-Based Channel Estimation for Wireless ultraviolet MIMO Communication Systems
Abstract:
To solve the problems of pulse broadening and channel fading caused by atmospheric scattering and turbulence, multiple-input multiple-output (MIMO) technology is an valid way. A wireless ultraviolet MIMO channel estimation approach based on deep learning is provided in this paper, deep learning is used to convert the channel estimation into the image processing. By combining convolutional neural network (CNN) and attention mechanism (AM), the learning model structure is designed to extract the depth features of channel state information (CSI). The simulation results show that the approach proposed in this paper can perform channel estimation effectively for ultraviolet MIMO communication and can better suppress the fading caused by scattering and turbulence in the MIMO scat-tering channel.
- 1
-
Analysis on influencing factors of detecting chemical oxygen demand in water by three-dimensional spectroscopy
Abstract:
This paper focuses on the standard COD liquid and studies the impact of pH, nitrite nitrogen, nitrate nitrogen, heavy met-als, salinity, and other factors on fluorescence intensity and fluorescence peak position during the detection of COD in water using fluorescence spectrometry. The influence mechanisms of different environmental factors on fluorescence spectra are also analyzed. Results indicate that pH value affects the fluorescence emission wavelength (Em), resulting in a redshift from 1.5 to 7.2, and a blueshift from 7.2 to 12.3. Nitrate nitrogen can react with organic matter in water to form nitro compounds, leading to a decrease in fluorescence intensity. Salinity has a negligible effect on T1 peak but a rela-tively large effect on T2 peak. Heavy metal ion concentration has a significant impact on T2 peak, while T1 peak position shifts with an increase in heavy metal ions. This study aims to explore the factors that can impact the detection of COD in water using three-dimensional fluorescence spectrometry, providing references to improve accuracy and practicability for COD detection based on three-dimensional fluorescence spectrometry.
- 1
-
Polarization-dependent ultrafast optical nonlinearities of N,N-dimethylmethanamide at 400 nm
Xin Zhao, Zhiyong Xue, Qingyang Liu, Xiaoqing Yan
Abstract:
Ultrafast optical nonlinearities of N,N-dimethylmethanamide (DMF) are studied by using polarized light at 400 nm. Both nonlinear refraction and stimulated Rayleigh-wing scattering (SRWS) depend on the polarization state of incident beam, while two-photon absorption changes negligibly with polarization state. The polarization dependence of SRWS originates from that of nonlinear refraction via self-focusing effect. Third-order susceptibility elements of DMF were determined, and a method to distinguish the multi-photon absorption signal from SRWS in Z-scan is provided. These results are helpful for the nonlinear optical research of the novel materials dissolved in DMF.
- 1
-
Research on multi-parametric sensors based multi-mode microfiber
Abstract:
A multiparameter sensor based on multimode microfiber is proposed, utilizing the modal interference between HE11 mode and HE12 mode in the elongated multimode fiber to achieve the temperature and pressure measure-ment. In this paper, the simulation model of modal interference based on multimode microfiber is established and the mechanism of modal interference is analyzed. Using the different mechanisms of modal response in the fiber at different wavelengths, the temperature was inverted using the offset of wavelengths in the spectrum, and the pressure was measured using the change of light intensity. The independent measurement of temperature and pressure was achieved. The experimental results show that the sensor has a temperature sensitivity of 1.4 nm/?C. In the case of pressure sensing, the sensor shows a sensitivity of -0.16 dBm/g.
- 1
-
Vehicle and pedestrian detection method based on improved YOLOv4-tiny
Abstract:
Aiming at the problem of low detection accuracy of vehicle and pedestrian detection models, this paper proposes an im-proved YOLOv4-tiny vehicle and pedestrian target detection algorithm. Attention Module (CBAM) is introduced into CSPDarknet53-tin module to enhance feature extraction capabilities; In addition, the CSP-DBL module is used to replace the original simple convolutional module superposi-tion, which compensates for the high-resolution characteristic information and further improves the detection accuracy of the network. Finally, the test results on the BDD100K traffic dataset show that the mAP value of the final network of the proposed method is 88.74%, and the detection speed reaches 63FPS, which improves the detection accuracy of the network and meets the real-time detection speed.
- 1
-
A deep learning based fine-grained classification algorithm for grading of visual impairment in cataract patients
jiangjiewei, zhangyi, xiehe, gongjiamin, lizhongwen
Abstract:
Recent advancements in artificial intelligence (AI) have shown promising potential for the automated screening and grading of cataracts. However, the different types of visual impairment caused by cataracts exhibit similar pheno-types, posing significant challenges for accurately assessing the severity of visual impairment. To address this issue, we propose a dense convolution combined with attention mechanism and multi-level classifier (DAMC_Net) for visual impairment grading. First, the double attention mechanism is utilized to enable the DAMC_Net to focus on lesions-related regions. Then, a hierarchical multi-level classifier is constructed to enhance the recognition ability in distinguishing the severities of visual impairment while maintaining a better screening rate for normal samples. In addition, a cost-sensitive method is applied to address the problem of higher false-negative rate caused by the im-balanced dataset. Experimental results demonstrated that the DAMC_Net outperformed ResNet50 and Dense-Net121 models, with sensitivity improvements of 6.0% and 3.4% on the category of mild visual impairment caused by cataracts (MVICC), and 2.1% and 4.3% on the category of moderate to severe visual impairment caused by cataracts (MSVICC), respectively. The comparable performance on two external test datasets was achieved, further verifying the effectiveness and generalizability of the DAMC_Net.
- 1
-
Dual-band reflective polarization converter based on metasurface
Lin Xiaofang, Zhang Xu, Chang Ming, Yu Siyang
Abstract:
In this paper, a dual-band and reflective polarization converter based on metasurface is proposed. Its unit cell is composed of two layers of metal plates separated by a dielectric substrate. The simulation results show that the proposed converter is able to convert x- or y-polarized incident waves into cross-polarized waves perfectly in frequency bands of 6.75-10.59 GHz and 17.78-19.61GHz,and the polarization conversion ratio is nearly 100%, which can also convert the linearly polarized waves into circularly polarized waves. It’s rarely to realize polarization conversion simultaneously, which can be widely used in radar satellites, antenna design and telecommunication applications.
- 1
-
Electrical capacitance tomography image reconstruction based on iterative Tikhonov regularization improved algorithm
Abstract:
Aiming at the problems of low reconstruction quality, poor robustness, and the inability to quickly and stably converge caused by the ill-posedness of electrical capacitance tomography image reconstruction, an improved algorithm based on iterative Tikhonov regularization was proposed. The algorithm constructs a new objective function by introducing the norm to carry out multi-criteria constraints, and introduces the result of the corrected Tikhonov regularization algorithm into the image reconstruction process together with the logarithmic weight factor as the estimated value. At the same time, an acceleration strategy is used, and the residual term is exponentially filtered. Perform ablation, initial value sensitivity, convergence, and noise interference experiments on the improved algorithm and compare it with other common algorithms. Experimental results show that the improved algorithm can quickly and stably converge and has good robustness and initial value insensitivity. The reconstructed image quality is high, the average correlation coefficient can reach 0.9633, and the average relative error can be reduced to 0.0694.
- 1
-
Properties of sandwich-structure based Mach-Zehnder Interferometer Cascade with FBGs Response to Dual-parameter Sensing
Abstract:
An all-fiber temperature and curvature sensor based on Mach-Zehnder interferometer (MZI) was proposed. The MZI was a sandwich structure which is composed by ring-core fiber (RCF), no-core fiber (NCF) and single-mode fiber (SMF). The temperature and curvature can be demodulated by the wavelength shift and the intensity variation of the dips respectively in the transmission spectrum. The measurement results show that the sensitivity of curvature is -7.88 nm/m-1 in the range from 3.0 m-1 to 4.2 m-1 and temperature is 53.5 pm/oC in the range from 60 oC to 200 oC. In addition, the cascaded FBG in the proposed structure, also sensitive to temperature, was used to monitor the fluctuation of temperature. The compact structure, the real-time temperature and the high curvature sensitivity make the sensor has the potential in the field of construction health monitoring and mining safety production.
- 1
-
Density-matrix-formalism based scheme for polarization mode dispersion monitoring and compensation in optical fiber communication systems
Du Qiuping, Zhang Xia, Guo Yao, Yang Zhenshan, Zhang Xiaoguang
Abstract:
We propose a density-matrix-formalism based scheme to study polarization mode dispersion (PMD) monitoring and compensation in optical fiber communication systems. Compared to traditional monitoring and compensation schemes based on the PMD vector in the Stokes space, the scheme we proposed requires no auxiliary matrices and can be handily extended to any higher-dimensional modal spaces, which is advantageous in mode-division multi-plexing (MDM) systems. A 28GBaud polarization division multiplexing quadrature phase-shift keying (PDM-QPSK) coherent simulation system is built to demonstrate that our scheme can implement the monitoring and compen-sation of 170 ps large differential-group-delay (DGD) that far exceeds the typical DGDs in practical optical com-munication systems. The results verify the effectiveness of the density-matrix-formalism based scheme in PMD monitoring and compensation, thus pave the way for further applications of the scheme in more general MDM optical communication systems.
- 1
-
Development of a Bias Power Supply for Geiger Mode Avalanche Photodiodes
Abstract:
Avalanche photodiodes have high output and high stability requirements for bias power in Geiger mode. This paper designs an avalanche photodiode (APD) with high boost ratio, high precision, low temperature drift, small size, and low power. Bias power supply, this module uses switching chip IC and flyback transformer to achieve high step-up ratio, realizes precise output control through precision op amp and T-type resistor feedback network, and designs appropriate compensation network to improve system stability. The size of the module is 2.5?2.5cm, the output voltage is adjustable from 0 to 450V, and the maximum ripple does not exceed 5.4mV. By changing the control voltage, any type of APD in Geiger mode can be biased, and the maximum deviation of the bias voltage does not exceed 0.5%.
- 1
-
Finite element simulation of Rayleigh surface acoustic wave in (100)AlN/(100)ZnO/diamond layered structure *
Ze-yu Zhang, Jin Qian, Li-rong Qian, Fu-jun Wen, Li-tian Wang, Cui-ping Li
Abstract:
(100)AlN/(100)ZnO/diamond layered structures are theoretically simulated by finite element method (FEM) to investigate the Rayleigh surface acoustic wave (SAW) propagation properties, including phase velocity, electromechanical coupling coefficient K2, and temperature coefficient of frequency (TCF). Three types of layered structures with different IDTs arrangements, which are IDTs/(100)AlN/(100)ZnO/diamond, (100)AlN/IDTs/(100)ZnO/diamond, and (100)AlN/(100)ZnO/IDTs/diamond structures, are considered in the simulations. The results show that the Sezawa mode exhibits larger K2 than the other modes. We found that the (100)AlN/IDTs/(100)ZnO/diamond structure exhibited better SAW properties including high K2 and appropriate phase velocity.
- 1
-
Silicon-based Fano resonance devices based on photonic crystal nanobeams
WANG Yihao, LU Wenda, LAI Xiaohan, DONG Mingli, LU Lidan, ZHU Lianqing
Abstract:
To address the driving power and density of wavelength-division-multiplexing (WDM) computing architectures, a Fano resonator based on a photonic crystal nanobeam is proposed. The Fano resonator comprises a T-shaped waveguide, introducing an additional phase shift in the continuous propagation mode, and a photonic crystal nanobeam with a discrete mode. The device has one resonance peak within wavelength ranging from 1500 nm to 1600 nm, with a maximum extinction ratio of 8.7 dB and a transmission spectrum slope of up to 11.30 dB/nm. The device has good reusability, extinction ratio, and spectral resolution. It is expected to provide essential photonic components for low-energy consumption and high-density photonic computing to meet the requirements of future convolutional neural network acceleration computing.
- 1
-
Ultra-large mode area multi-core orbital angular mo-mentum transmission fiber designed by neural network and optimization algorithms
Abstract:
A large mode area multi-core orbital angular momentum (OAM) transmission fiber is designed and optimized by neural network and optimization algorithms. The neural network model has been established first to predict the optical properties of multi-core OAM transmission fibers with high accuracy and speed, including mode area, nonlinear coefficient, purity, dispersion, and effective index difference. Then the trained neural network model is combined with different particle swarm optimization algorithms for automatic iterative optimization of multi-core structures respectively. Due to the struc-tural advantages of multi-core fiber and the automatic optimization process, we designed a number of multi-core struc-tures with high OAM mode purity (>95%) and ultra-large mode area (>3000μm2), which is larger by more than an order of magnitude compared to the conventional ring-core OAM transmission fibers.
- 1
-
A high dynamic range pixel using lateral overflow integration capacitor and adaptive feedback structure in CMOS image sensors
YUEJIN YANG, JIANGTAO XU, BIAO MA, QUANMIN CHEN, KAIMING NIE
Abstract:
This letter proposes a novel high dynamic range (HDR) pixel using lateral overflow integration capacitor (LOFIC) and adaptive feedback structure. Through detailed analysis of the voltage feedback mechanism, the conversion gain (CG), full well capacity (FWC) and DR performances of the feedback LOFIC pixel are analytically expressed. The verification results reveal that the equivalent FWC of the feedback LOFIC pixel is 1.89 times of conventional LOFIC pixel, and the DR extension is 5.5 dB. Based on 110 nm CMOS process, a 5.0 μm pixel layout is presented, using 13.3 fF capacitance to achieve 83 ke- FWC and 102.8 dB DR, which are 44 ke- and 97.3 dB of conventional LOFIC pixel under the same design conditions. This also demonstrates that the feedback LOFIC pixel can reduce the de-pendence of extended DR on capacitor area, and can be used as a reference for HDR pixels design.
- 1
-
Double Fano resonances in disk-nonconcentric ring plasmonic nanostructures
Abstract:
The plasmonic properties of gold nanostructures composed of a disk outside a nonconcentric ring are numerically studied by the finite difference time domain method. Simulated results show that two Fano resonances are formed as a result of the coupling of the octupolar and quadrupolar modes of the ring with the dipolar mode of the disk. The reduction in structural symmetry causes a red shift of the Fano resonances and distinct changes in spectral lineshape by offsetting the center of the inner surface of the ring to different directions. The effects of several geometric parameters on the characteristics of Fano resonances are also discussed. In addition, the refractive index sensitivities for the two Fano resonances can be up to 581 nm/RIU and 780 nm/RIU with the corresponding figure of merits as large as 12.7 and 10.2, respectively. Such properties render the structures useful for potential applications in multiwavelength sensors.
- 1
-
An adaptive graph embedding method for feature ex-traction of hyperspectral images based on approximate NMR model
Hong Qiu, Rengfang Wang, Heng Jin, Feng Wang
Abstract:
This paper introduces ANMRP, an adaptive graph embedding method for feature extraction of hyperspectral images. ANMRP utilizes an approximate NMR model to construct an adaptive neighborhood map between samples. The glob-ally optimal weight matrix is obtained by optimizing the approximate NMR model using fast ADMM. The optimal projection matrix is then determined by maximizing the ratio of the local scatter matrix to the total scatter matrix, al-lowing for the extraction of discriminative features. Experimental results demonstrate the effectiveness of ANMRP compared to related methods.
- 1
-
Fabrication and characterization of high-damage resistance Zn-diffused MgO:PPLN ridge waveguides
Abstract:
This study investigates the fabrication process of zinc-diffused ridge waveguides in periodically poled magnesi-um-doped lithium niobate (PPMgO:LN). A controlled variable method is used to study the effects of diffusion tem-perature, diffusion time, ZnO film thickness, and barrier layer thickness on the surface domain depolarization and waveguide quality of PPMgO:LN. A special barrier layer is proposed that can automatically lift off from the sample surface, which increases the depth of zinc doping and reduces the surface loss of the waveguide. By optimizing the process parameters, we fabricate zinc-diffused PPMgO:LN ridge waveguides with a length of 22.80 mm and a period of 18.0 μm. The above waveguides generate a second harmonic (SHG) at 775 nm with an output power of 90.20 mW by a pump power of 741 mW at 1550 nm. The corresponding conversion efficiency is 3.160%/W∙cm2, and the wave-guide loss is approximately 0.81 dB/cm. These results demonstrate that high-efficiency devices can be obtained through the fabrication process described in this paper.
- 1
-
Compact yellow-orange Nd:YVO4 / PPMgLN laser at 589nm
Lei Ma, Feng xinkai, Chen huaixi, Chen jiaying, cheng xing, Liang wanguo
Abstract:
In the paper, we propose and make a compact yellow-orange laser of the Nd:YVO4/PPMgLN module by Raman frequency-doubling at 589 nm for the first time. By reasonably designing the size of the Nd-doped Yttrium Vanadate (Nd:YVO4) and 5 mol% Mg-doped periodically poled lithium niobate (PPMgLN) crystals, cavity length and coating parameter, a compact 589 nm laser module with a total size of 3 *10 *1.5 mm3 is fabricated. In the laser module, the input surface of Nd:YVO4 crystal is end-pumped by an 808 nm laser diode (LD). Under the effect of linear resonant cavity structure, the output surface of PPMgLN crystal with a period of 9.48 μm generates 589 nm yellow-orange light. The experimental results show that the maximum output power at 589 nm is 390 mW at the pump power of 3 W with the optical-optical conversion efficiency of 13% and the stabililty of the output power is less than 2% within 3 h.
- 1
-
Improved Forward Error Correction Technology of RS-LDPC Cascade Code in Optical Transmission Network
Ma Xiurong, Ma Xiaohang, Shan Yunlong
Abstract:
With the continuous development of optical communication and the increase in data transmission volume, optical transmission networks (OTN) have become the focus of research on next-generation transmission networks. In the process of data transmission, errors caused by noise often occur, resulting in an increase in the bit error rate and a decrease in the performance of the optical communication system. Therefore, we use forward error correction technology (FEC) in OTN for error control to improve the transmission efficiency of signals in OTN and reduce the bit error rate.Standard FEC technology uses RS(255,239) codes. On this basis, since the performance of LDPC codes is close to the Shannon limit, we propose a method of cascading RS codes and LDPC codes.Applying this improved FEC technology to OTN, the simulation results show that the improved FEC technology has a reduced bit error rate compared with the standard FEC technology. When the bit error rate is the minus 3th power of 10, the performance is improved by about 1.7 dB.
- 1
-
An automatic docking method for large-scale sections based on real-time pose measuring and assembly deviation control
Qiao Zhifeng, Fu Kang, Liu Zhenzhong
Abstract:
Aiming at the problem of poor accuracy consistency of large sections’ docking assembly, an automatic docking method using multiple laser trackers to measure the position and posture of the docking sections in real time was proposed. In the solution of the pose of the docking section, real-time pose measurement of the docking section was realized by establishing a global coordinate system and a coordinate fusion method of three or more laser trackers. In the automatic control of the docking process, the real-time communication protocol and the circular negative feedback control strategy of measurement-adjustment-remeasurement are adopted, and the fully-automated docking of large sections is realized. Finally, an experimental verification system was set up, and the docking of the large-scale section reduction models was realized under the requirements of docking accuracy, and the effectiveness of the automatic docking scheme was successfully verified.
- 1
-
A lightweight global awareness deep network model for flame and smoke detection
Abstract:
Aiming at the trouble of low detection accuracy and the problem of large model size, this paper proposes a lightweight flame-and-smoke detection model depending on global awareness of images, named GAL-YOLOv5. The proposed method replaces the CBS module of original YOLOv5 in the backbone with DBS, and the C3 module with GC3, and thus constructs a lightweight backbone network DGNet. Besides, Involution (InvC3) module is proposed to enhance the global modeling ability and compress the model size, and a module using adaptive receptive fields, named FConv, is proposed to enhance the model’s perception capacity for foreground complex flame-and-smoke information in feature maps. Experimental results show that GAL-YOLOv5 increases mAP@0.5 to 70.8%, mAP@0.5:0.95 to 39.7%, reduces the number of parameters to 3.57M and the amount of calculation to 7.4GFLOPs under the premise of ensuring the detection speed. It has been verified that the model can achieve high-precision real-time detection of flame and smoke.
- 1
-
Availability Evaluation Model for Space-based Optical Aerial Target Detection System
Zhao Jiaqing, Zhang Lei, Ding Xiang, Xu Zhongchao, Fan Shiwei, Liu Rongke
Abstract:
For space-based optical detection systems, there is usually a difference between actual on-orbit operational perfor-mance and designed requirements based on fixed scenarios. To assess the availability of space-based optical detection systems in different backgrounds, the radiation characteristics of aerial targets have been simulated using body radia-tion and atmospheric transmission models. The background radiation characteristics were also statistically analyzed. Then, for the parameters of the fixed space-based optical detection system, the signal-to-clutter and availability were evaluated under different conditions. A linear relationship between the radiation intensity and the flight height of the target was obtained. For a space-based optical detection system, the analytical availability model was constructed. Fi-nally, multiple groups of data under different simulation conditions were used to validate the universality and reliabil-ity of the model. This availability model could significantly reduce the time required to predict the availability of the space-based optical detection system. The model was also adopted to analyze the influence of flight height, mean and variance, and background clutter on the space-based optical detection availability.
- 1
-
Energy transmittance of focused femtosecond pulses at different air pressures
Suyu Li, Yu Miao, Zhang He, Cai Xiaoming, Jin Mingxing, Wu Jiabin
Abstract:
Transmission of intense femtosecond laser pulses in air is accompanied by energy depletion. By measuring the transmitted spectra of the focused femtosecond pulses in air, we study the influence of air pressure and initial pulse energy on the spectra around the central wavelength (800 nm) after the interaction of the focused femtosecond laser with air. On this basis, the energy transmittance of the central wavelength of the femtosecond pulses is obtained. It is found that as the air pressure is lower than 1 kPa, the transmitted spectra of focused femtosecond pulses change with the pressure, but there is almost no energy depletion; while as the air pressure is higher than 1 kPa, femtosecond filamentation occurs and the energy transmittance of the central wavelength of the femtosecond pulses decreases with the increase of air pressure and pulse energy. According to the different regimes (i.e., nonfilamentation, and filamentation regimes), we discuss the effect of energy conversion and transfer on the energy transmittance. This work can help to understand the energy depletion during the transmission of ultrashort intense laser pulses in air and provide a guidance for the practical applications of femtosecond filamentation.
- 1
-
Bifurcated convolutional network for Specular Highlight Removal
Xu Jingting, Liu Sheng, Chen Guanzhou, Liu Qianxi
Abstract:
Specular highlight usually causes serious information degradation, which leads to the failure of many computer vision algorithms. We have proposed a novel bifurcated convolution neural network (Bifurcated-CNN) to tackle the problem of high reflectivity image information degradation. 1) The specular highlight features are extracted and removed in two stages from coarse to fine, to ensure the generation of diffuse images have no visual artifacts and information distortions. 2) A bifurcated feature selection strategy (BFSS) is designed to filter out the specular highlight features and enhance the detection capability of our network. The experiments on two types of challenging datasets demonstrate that our method outperforms state-of-the-art approaches for specular highlight detection and removal. The effectiveness of the proposed BFSS and Bifurcated-CNN are also verified.
- 1
-
Exploring the growth mechanism of CuSbSe2 thin film prepared by electrodeposition
Abstract:
Copper antimony selenium (CuSbSe2), has advantages of adjustable band gaps from 1.09 to 1.2 eV, high light absorption coefficient (>105 cm-1), and low grain generation temperature (300~400 ℃), which is suitable for the preparation of solar cells. However, the stable range of CuSbSe2 (CASe) phase is narrow, which is inevitable to form Sb2Se3 and Cu3SbSe4 second phase during the preparation process. In this work, selenization annealing of Sb/Cu metal layer to prepare CASe thin films with pulse electrodeposition process was studied, and the growth mechanism of CASe film was analyzed. Cu and Sb reacted with Se to form Cu2Se and Sb2Se3, respectively. Then Cu2Se and Sb2Se3 further reacted to generate CASe. Since the formation temperature of Cu3SbSe4 was lower than that of CASe, the preferential formation of Cu3SbSe4 led to layer separation. When the annealing temperature was too high, CASe decomposed to form Cu3SbSe3 and Sb2Se3. Additionally, by increasing the heating rate, the separation of CASe thin films was effectively improved, and the CASe thin films with relatively high crystallinity were obtained at 360 ℃ with heating rate of 30 ℃/min and selenization time of 20 min.
- 1
-
Fabricating lifted Haar transform image compression optical chip based on femtosecond laser
Qing Tao, Liangpeng Wei, Wenxiang Kuang, Yegang Yin, Jian Cheng, Dun Liu
Abstract:
In this paper, a lifted Haar transform image compression optical chip has been researched to achieve rapid image compression. The chip comprises 32 same image compression optical circuits, each circuit contains a 2 × 2 multimode interference coupler and a π/2 delay line phase shifter as the key components. The chip uses highly borosilicate glass as the substrate, Su8 negative photoresist as the core layer, and air as the cladding layer. Its horizontal and longitudinal dimensions are 8011 μm × 10000 μm. Simulation results present that the designed optical circuit has a coupling ratio of 0:100 and an insertion loss of 0.001548 dB. Then the chip is fabricated by femtosecond laser and testing results illustrate that the chip has a coupling ratio of 6:94 and an insertion loss of 0.518 dB. So, the prepared chip possesses good image compression performance.
- 1
-
Obstacle Detection：Improved YOLOX-S based on Swin Transformer-Tiny
Abstract:
To address the accuracy challenge in obstacle detection for autonomous driving, we propose an improved YOLOX-S ob-stacle detection model that can detect multiple targets, including people, cars, bicycles, motorcycles, and buses. This paper aims to propose a model based on YOLOX-S that surpasses the baseline model and is capable of real-time detection:(1) We suggest that the existing YOLOX-S backbone be replaced with the Swin Transformer-Tiny backbone. This change aims to improve the local feature extraction capability, leading to more accurate detection of obstacles under real-world vehicle conditions. (2)We decreased the number of channels between the Swin Transformer and PA-FPN from [96, 192, 384, 768] to [192, 384, 768]. This reduces the computational cost and makes the Swin Transformer-Tiny more compati-ble with the PA-FPN. Conclusively, compared to YOLOX-S, our proposed method, YOLOX-S based on swin trans-former-tiny (ST-YOLOX-S), has a 6.1% improvement in accuracy on the COCO dataset. Among the five types of obsta-cles that will appear in simulated actual vehicle conditions, our ST-YOLOX-S has shown excellent improvement in accu-racy compared to YOLOX-S. Furthermore, the detection accuracy is significantly improved compared to YOLOv3, showing the effectiveness of the proposed algorithm.
- 1
-
Phase unwrapping based on deep learning in light field fringe projection 3D measurement
Abstract:
Phase unwrapping is one of the key roles in fringe projection three-dimensional (3D) measurement technology. In this paper we propose a new method to phase unwrapping in camera array light filed camera fringe projection 3D measurement base on deep learning. A multi-stream convolutional neural network is proposed to learn the mapping relationship between camera array light filed wrapped phases and fringe orders of the expected central view, and is used to predict the fringe order to achieve the phase unwrapping. Experiments are performed on the light field fringe projection data generated by simulated camera array fringe projection measurement system in Blender and by experimental 3*3 camera array light field fringe projection system. The performance of the proposed network with light field wrapped phases using multi directions as network input data is studied, and the advantages of phase unwrapping based on deep learning in light filed fringe projection are demonstrated.
- 1
-
Tuning the Photocatalytic Properties of TiO2-doped Single-walled Boron-nitride Nanotubes (SWBNNT) for Overall Water Splitting: A DFT Studies
Abstract:
This research investigated the potentials of the single walled boron-nitride nanotube doped with transition metal dichalcogenides as photocatalyst in hydrogen fuel production via overall water splitting technique. Studies of the electronic and optical absorption spectra analysis were done using the popular density functional theory and the time-dependent density functional theory. TiO2-doped SWBNNT photocatalysts was found with higher water oxidation (OH-/O2) and reduction potentials (H /H2) which agrees with the obtained literature values for most of the photocatalysts. Moreover, TiO2-doped SWBNNT demonstrated an indirect band of 2.7 eV and absorbs photon in the visible region of the electromagnetic spectrum. Studies of the optical absorption potential revealed that TiO2-doped SWBNNT absorbs and emits heat energy simultaneously, its absorption efficiencies are observed to be higher than other photocatalysts reported. Based on this behaviors and other results obtained, TiO2-doped SWBNNT can serve excellently as the photocatalyst for hydrogen evolution via water splitting technique.
- 1
-
REALIZATION OF 16 GB/S ALL-OPTICAL TOGGLE MEMORY UTILIZING CHANGE IN POLARIZATION STATE OF LIGHT IN BIDIRECTIONAL OPTICAL FIBER
Abstract:
In this investigation all-optical Toggle flip-flop event-driven memory is explored with data rate of 16 Gb/s. Bidirectional optical fiber is used as a nonlinear medium to generate the output set and reset pulses of a toggle flip-flop, the model is built using the bidirectional optical principle, taking into account the fundamental effects of cross phase modulation and self-phase modulation with change in polarization state. The performance of a flip-flop is evaluated using truth table conditions and performance parameters such as Q factor, which is obtained as 380.92 dB for Q and 272. 9 dB for Q?, rising and falling times of 7.304 ps and 5.79 ps, respectively, is obtained which makes flip-flop design fast.
- 1
-
A method for online detection and lifespan evaluation of light sources
Abstract:
LED light source degradation detection and lifetime evaluation usually use the data of light flux change as a basis, but the process of light flux measurement is complicated and tedious, requiring the use of an integrating sphere, and cannot be performed online. This is unfriendly to the detection of machine vision light sources used in production lines. To address this problem, this paper proposes and designs a method for online detection and lifetime evaluation of light sources by using a mini spectrometer to detect the intensity of light sources online, and evaluates the light degradation and lifetime of light sources based on the changes in light intensity during use. This determines whether the light source needs to be adjusted or replaced, avoiding misjudgment or missed judgment in production detection due to light source degradation. The experiment was conducted on LED under high-temperature accelerated aging. The light intensity data after aging was fitted by an acceleration and life evaluation model, and the fitting result showed that the error of life evaluation by this method was 8.37%. As a comparison, this paper also detects the changes in light flux of light sources during the experiment, and the average error between the decay of light intensity and the decay of light flux was only 0.102%. It has been validated that the error in evaluating the lifetime of light sources using this method is 10.71%, and the accuracy of the evaluation is about 90%. The results show that the method is accurate, reliable, and can be used as a basis for online detection and evaluation of LED light source degradation.
- 1
-
Single mode fiber dislocation Mach–Zehnder interferometer cascaded with Fiber Bragg grating for monitoring of metal electrochemical corrosion
zhangshuaibo, liuxiaoqi, wangzhi, liuyange
Abstract:
In this paper, an optical fiber sensor composed of a dislocation Mach–Zehnder interferometer (MZI) cascade with a fiber Bragg grating (FBG) is proposed, and it used to monitor the electrochemical corrosion of metals in experiments. The dislocation interferometer is composed of two segments of single-mode fiber and one segment of dislocation single-mode fiber. The contact surface is increased between the fiber and the environment which helps to improve the interference sensitivity. The relationship between the dislocation amount and the refractive index sensitivity of the interferometer is discussed through simulation. In the experiment, the sensitivity of the interferometer reaches more than 10000 nm/RIU, and the monitoring of metal electrochemical corrosion is also realized in 3.5% NaCl solution. The proposed measurement scheme has the advantages of small structure, low cost and high sensitivity. It has good prospects in chemical reaction monitoring.
- 1
-
Fast image reconstruction network in image stitching
Abstract:
Compared with the traditional feature-based image stitching algorithm, the free-view image stitching algorithm based on deep learning has the advantages of fast stitching speed and good effect. However, these algorithms still cannot achieve real-time splicing speed. For the image reconstruction stage, we redesign a new fast image reconstruction network. This network is designed based on ShuffleNet, and the new network structure and loss function will reduce the time required for image reconstruction. In addition, this network can also reduce the performance loss after the network is lightweight. It is proved by experiments that the fast image reconstruction network can realize real-time high-resolution image reconstruction.
- 1
-
Thermal radiation effect in near infrared single photon detector
Abstract:
A miniaturized near infrared single photon detector is demonstrated by integrating a thermoelectric cooler, a thermistor, and a planar type InGaAs/InP separate absorption, grading, charge and multiplication structure single photon avalanche diode into a butterfly case. The performance of the device at different temperatures is tested. It can achieve 20.3% single photon detection efficiency and 1.38 kHz dark count rate when the chip is cooled to 223 K. The test results show that even when the chip temperature is kept constant, the dark count rate of the device still increases with the increase of the ambient temperature, which is inconsistent with the carrier generation mechanism of semiconductor materials. The mechanism is researched and it is found that the thermal radiation of the high temperature case is the main source of dark count.
- 1
-
Memory-Boosting RNN with Dynamic Graph for Event-based Action Recognition
Chen Guanzhou, Liu Sheng, Xu Jingting
Abstract:
Owing to its characteristic of high dynamic, event camera is well suited for capturing subtle temporal changes in action recognition applications. However, existing action recognition methods based on event cameras have not fully exploited the advantages of event cameras, such as compressing event streams into frames for subsequent calculation, which greatly sacrifices the time information of event streams. Meanwhile, the conventional PointCloud-based methods suffer from large computational complexity while processing event data, which make it difficult to handle long-term actions. To tackle the above problems, we propose a Recurrent Neural Network with Memory-Boosting and Dynamic Graph(DG-MBRNN). The proposed DG-MBRNN splits the event stream into sequential graph data for preserving structural information, then uses the RNN with boosting spatiotemporal memory to handle long-term sequences of actions. In addition, the proposed method introduces a dynamic reorganization mechanism for the graph based on the distances of features, which can effectively increase the ability to extract local features. In addition, in order to cope with the situation that the existing data sets have too simple actions and too limited categories, we propose a new Event-based dataset containing 36 complex actions. This dataset will greatly promote the development of Event-based action recognition research. Experimental results show the effectiveness of the proposed method in completing the Event-based action recognition task.
- 1
-
Wavefront detection performance analysis of plenoptic sensor
jiangtao, zhangjinghui, wanghaitao, qiaochunhong, fanchengyu
Abstract:
A numerical simulation model of plenoptic sensor aberration wavefront detection is established to simulate and analyze the detection performance of plenoptic sensor aberration wavefront for different turbulence intensities. The results show that the plenoptic sensor can achieve better distortion wavefront detection, and its wavefront detection accuracy improves with turbulence intensity. The unique optical structure design of the plenoptic sensor makes it more suitable for aberration wavefront detection in strong turbulent conditions. The wavefront detection performance of the plenoptic sensor is not only related to its wavefront reconstruction algorithm but also closely related to its structural parameter settings. The influence of structural parameters on the wavefront detection accuracy of plenoptic sensors under different turbulence intensities is simulated and analyzed. The variation law of wavefront detection accuracy and structural parameters under different turbulence intensities is summarized to provide a reference for the structural design and parameter optimization of plenoptic sensors.
- 1
-
Highly efficient convolution computing architecture based on silicon photonic Fano resonance devices
JiaRong Ni, 卢文达, LAI Xiaohan, LU Lidan, OU Jianzhen, ZHU Lianqing
Abstract:
Convolutional neural networks require a lot of multiplication and addition operations completed by traditional electrical multipliers, leading to high power consumption and limited speed. Here, a silicon waveguide-based wavelength division multiplexing architecture for convolution neural network is optimized with high energy efficiency Fano resonator. Coupling of T-waveguide and Micro-ring resonator generate Fano resonance with small half-width, which can significantly reduce the modulator power consumption. National Grid's insulator dataset is used to test Fano resonance modulator-based convolutional neural networks. The results show that accuracy for insulator defect recognition reaches 99.27% with much lower power consumption. Obviously, Our opti-mized photonic integration architecture for convolutional neural networks has broad potential for the artificial intelligence hardware platform.
- 1
-
Mid-wavelength InAs/GaSb Type-II Superlattice Barrier Detector with nBn Design and M Barrier
Liu Zhaojun, Zhu Lianqing, Lu Lidan, Dong Mingli, Zhang Dongliang, Zheng Xiantong, Liu Yuan, Yu Jing
Abstract:
This study reports the performance of an InAs/GaSb type-II superlattices (T2SLs) detector with nBn structure for mid-wavelength infrared detection. An electronic band structure of M structure is calculated using 8-band k?p method, and the nBn structure is designed with the M-barrier. Mesa isolation of the nBn structure and the removal of Al2O3 pas-sivation layer on the mesa are prepared by wet etch, which is simple in manufacturing process. X-ray diffraction (XRD) and atomic force microscope (AFM) characterization indicate that the detector material has good crystal quality and sur-face morphology. Spectral response measurements at different biases and temperatures are performed. The saturation bias is 300 mV, and the device is promising to work at a temperature of 140 K. Energy gap of T2SL versus temperatures are fitted by the Varshni curve, and zero temperature bandgap Eg(0), empirical coefficients α and β are extracted. A dark cur-rent density of 3.2?10-5 A?cm2 and differential resistance area (RA) product of 1.0 ? 104 Ω?cm2 are measured for the nBn detector at 77 K. The dominant mechanism of dark current at different temperature ranges is analyzed by calculating the activation energy according to the Arrhenius plot. The device with a 50% cutoff wavelength of 4.68 ?m exhibits a responsivity of 0.6 A/W without antireflection coating, a topside illuminated quantum efficiency of 20%, and a detectivity of 9.17?1011 cm?Hz1/2/W at 77 K and 0.3 V. Excellent performance indicates that the device is a promising candidate for large area focal plane arrays.
- 1
-
Modeling and analysis of actively Q-switched Fe:ZnSe laser pumped by a 2.8 μm fiber laser
Abstract:
A theoretical model concerning active Q-switching of an Fe:ZnSe laser pumped by a continuous-wave (CW) 2.8 μm fiber laser is developed. Calculations are compared with the recently reported experiment results, and good agreement is achieved. Effects of pump power, output reflectivity, ion concentration and temperature of crystal on the laser output performance are investigated and analyzed. Numerical results demonstrate that, similar to highly efficient CW Fe:ZnSe laser, low temperature of the crystal is significant to obtain high peak power Q-switched pulses. The numerical simulation results are useful for optimizing the design of actively Q-switched Fe:ZnSe laser.
- 1
-
Measurement of Absolute Three-dimensional Displace-ment Gradients Using Tri-color Quad-beam Digital Shearography
LI Xuan, WU Sijin, TANG Xiaojun, LI Weixian
Abstract:
A tri-color quad-beam digital shearography is proposed to achieve the measurement of absolute three-dimensional dis-placement gradients. Four laser beams with three different center wavelengths are symmetrically irradiated to the object surface from the upper and lower left and right directions. Four phase maps are then extracted from the two inter-ferograms obtained from two shots. Based on these four phase maps, the absolute three-dimensional displacement gra-dients are determined. This means of absolute three-dimensional displacement gradient measurement effectively improves the measurement capability of digital shearography and expands its application range.
- 1
-
Simulations of ultra-high sensitivity RI sensor in tri-ple-core fiber with modified Vernier effect: Application to marine RI measurement
XIONG Lingyi, YU Yang-fei, DUAN Shao-xiang, LIU Bo, LIN Wei, YAO Yuan
Abstract:
In this Letter, we proposed an ultra-high sensitivity triple-core fiber refractive index (RI) sensor with a modified Vernier effect for marine RI measurement and demonstrated it by numerical simulation. This sensor composes a pair of parallel-ized spatial mode Mach-Zehnder interferometers, both of which are involved in sensing, but possess different interfering modes. By designing a Mach-Zehnder Interferometer refractive index fiber optic sensor based on Vernier effect in air, it is demonstrated that in the low refractive index such as air environment, only the modes involved in sensing interference are affected by the environment to generate Vernier effect. In the high refractive index marine environment, both sensing interferometer and reference interferometer need to be affected by the ambient refractive index to generate Vernier effect. The simulation results indicate that the proposed novel sensing structure can amplify its sensitivity from -15428 nm/RIU to -24857 nm/RIU in the marine environment.
- 1
-
Image Analysis Considering Textual Correlations Enables Accurate User Switching Tendency Prediction
Jianbin Wang, Shuyuan Shi, Xuna Wang, Jiahui Yu
Abstract:
Predicting likely-to-churn users employing surveys is a challenging task. Individuals with different personalities may make different choices in the same situation, so we introduced social media avatars that reflect the user's psycho-logical state when analyzing their churn tendency. In this paper, we propose a multimodal framework that jointly learns image and text features to establish correlations among users with low NPS scores and those likely to churn. We conducted experiments on actual data, and the results show that our proposed method can identify NPS-degraded users in advance, promoting the commercial development of the operator.
- 1
-
Current measurement method based on magneto-optic rotation effect
shuihuahsneg, jiangchunlei, rongyufei, dongtaiji
Abstract:
Aiming at the approximate measurement of magnetic rotation angle in optical current sensor based on light intensity de-tection mode, this paper proposes a current measurement method based on triangular constant transformation to recon-struct magnetic rotation angle, so as to avoid the large current measurement error caused by the approximate measurement of the magnetic rotation angle. By extracting the DC component and the AC component of the light intensity signal de-tected by the photoelectric detector, the sine signal containing the magnetic rotation angle is directly obtained by dividing the two components, and then the triangular identity transformation method is used to linearly demodulate the magnetic rotation angle and reconstruct the current waveform. The experimental results show that the relative error of current measurement does not exceed 1.40% in the current range of 0.05A ~ 0.50A, which is less than the approximate linear measurement method, and the magnetic rotation angle and the current have a good linear relationship.
- 1
-
8-HQCdCl2H2O as an organic Q-switcher in Erbium laser cavity
Sulaiman Wadi Harun, MUSTAFA MOHAMMED NAJM, MOHAMMED NAJM ABDULLAH, AYA A. ALMUKHTAR, Bilal Nizamani, BELAL AHMED HAMIDA, Moh Yasin
Abstract:
This letter demonstrated a Q-switched Erbium-doped fiber laser using an organic saturable absorber based on 8-HQCdCl2H2O material. The organic thin film was prepared using the casting process. The proposed Q-switched EDFL has a maximum repetition rate of 143 kHz, a minimum pulse duration of 1.85 ?s and highest pulse energy of 167 nJ. The Q-switched peak laser was at central wavelength of 1531 nm with a 3-dB bandwidth of 3.52 nm and power intensity of 2.64 dBm
- 1
-
SFR-Net: Sample-aware and feature refinement network for cross-domain micro-expression recognition
LIU Jing, JI Xinyu, WANG Mengmeng
Abstract:
Over the past several decades, micro-expression recognition (MER) has become a growing concern for scientific community. As the filming conditions vary from database to database, previous single-domain MER methods generally exhibit severe performance drop when applied to another database. To deal with this pressing problem, in this paper, a sample-aware and feature refinement network (SFR-Net) is proposed, which combines domain adap-tation with deep metric learning to extract intrinsic features of micro-expressions for accurate recognition. With the help of decoders, siamese networks increasingly refine shared features relevant to emotions while exclusive features irrelevant to emotions are gradually obtained by private networks. In order to achieve promising performance, we further design sample-aware loss to constrain the feature distribution in the high-dimensional feature space. Ex-perimental results show the proposed algorithm can effectively mitigate the diversity among different mi-cro-expression databases, and achieve better generalization performance compared with state-of-the-art methods.
- 1
-
Remote detection of alpha radiation source by optical method
Abstract:
In order to solve the problem that the traditional alpha radiation source detection technique is unable to achieve re-mote detection and is not suitable for detecting alpha radiation source with complex morphology, remote detection of alpha radiation source by measuring the fluorescence generated by the interaction between alpha particles and nitro-gen molecules is investigated in this paper. In dark condition, nitrogen fluorescence is collected by the optical system with an aperture of 50 mm and detected by a photomultiplier tube working in the photon counting mode. The alpha radiation source with total activity of 7.3 kBq and unit area activity of 49 Bq/cm2 can be detected within 30 s at a distance of 0.6 m. The detection limit of the alpha radiation source detection system proposed in this paper ranks among the top in the related research fields.
- 1
-
High-efficiency and Reduced Efficiency Roll-off of Top-emitting Organic Light-emitting Diodes base on the Tandem Structure with Photovoltaic Effect
Abstract:
Top-emitting organic light-emitting diodes (TEOLED) have attracted extensive attention for the high-brightness and flat-panel display. However, the efficiency roll-off at high-brightness is the issue that need to be resolved for further practical applications using TEOLED devices. Herein, a serials of high-efficiency tandem TEOLED introducing a fullerene/zinc-phthalocyanine organic semiconductor heterojunction as charge generation layer are demonstrated. With unique photovoltaic property, the charge generation layer can absorb part of the photons emitted by the emission layer (Ir(ppy)3) and generate electrons and holes. By optimizing the charge generation layer thickness, the as-fabricated TEOLED device shows a high luminance of 156000 cd/cm2 and pure green electroluminescence with a maximum current efficiency of 86 cd/A. Importantly, relying on the energy between the photovoltaic effect and the microcavity effect, only 1.5% of the efficiency roll-off is obtained at 1000-10000 cd/cm2. We believe the introduction of fullerene/zinc-phthalocyanine as charge generating layer here provide a very promising alternative way for developing high-efficiency tandem TEOLED devices.
- 1
-
Dual-Band THz Antenna Based on a Novel Photonic Band Gap Structure
Abstract:
In this paper, a novel dual-band Photonic Band Gap (PBG) structure was proposed and applied to a dual-band microstrip antenna. The antenna resonates at 1.267 THz and 1.502 THz, exhibiting reflection coefficients of -58.177, and -49.462 dB, and gains of 3.173, and 5.232 dBi, respectively. Compared with the microstrip antenna based on the homogeneous silicon substrate, the designed dual-band antenna based on the novel PBG structure shows improved impedance matching, radiation efficiency, and gain. The paper simulated and analyzed the impacts of different filling dielectric materials and the variations of dimensions, position, period, and corrugation depths of the du-al-band PBG structure on the resonances of the antenna. It is expected that the proposed novel dual-band PBG structure has great application potential，Such cancer diagnosis is done by utilizing the radiation characteristics of the antenna.
- 1
-
Lattice vibration, optical, and mechanical properties of Aluminum Phosphide (AlP) compound under the influence of temperature
Abstract:
The manuscript discusses the lattice vibrations, optical, and mechanical characteristics of the zinc-blende AlP compound. Investigations have been done into the effect of temperature on refractive index, optical dielectric constant, static dielectric constant, longitudinal and transversal sound velocities, reflectivity, susceptibility, phonon frequencies, micro-hardness, ionicity, and transverse effective charge. The calculations in this article were carried out using the pseudo-potential method (EPM). Comparative analysis with the existing experiment and other theoretical calculations reveals a satisfactory level of agreement. The optoelectronic applications in the high pressure region could make use of the predicted properties.
- 1
