最新刊期
期 21 ， 2023
- 摘要：In order to improve the dynamic range of LiDAR and to address the issue of off-target measurements in high-speed scanning LiDAR systems when capturing long-range targets， a method that involves deviation of the detector from the focal plane is proposed. The off-target phenomenon of distant targets， the principle of radial and axial deviation of the detector in the focal plane， and the spatial energy distribution of the echo spot near the focal plane are investigated. First， the scanning speed， focal length， and defocusing amount of the laser and the radial deviation angle and radial deviation amount of the echo spot center relative to the optical axis are analyzed according to the target distance. Next， the effect of variation in echo spot size on the detector is analyzed when the relative position relationship between the image plane， the focal plane， and the detector changes. Subsequently， the energy distribution of the echo spot in the detector and the light-receiving efficiency are evaluated， considering the effects of the energy distribution of the Gaussian beam， the aperture diaphragm of the optical system， diffraction， and defocusing. Finally， the variation law of the light-receiving efficiency of the target echo at different distances after the detector deviates from the focal plane is verified using numerical simulations and experiments. The experimental results indicate that the farthest target can be measured from approximately 800 to 1 300 m by adjusting the radial offset under the 75 Hz quadrilateral tower mirror-scanning mode. By adjusting the axial deviation， the light-receiving efficiency of the 5 m target can be decreased by approximately 70%， whereas that of the long-range target remains basically unchanged. The proposed method solves the problem of off-target echo in long-range targets and improves the dynamic range of LiDAR.关键词：lidar;Dynamic range;off-target;defocusing
- 58
- |
- 20
- |
- 0
发布时间：2023-11-16 - 摘要：The unpolarizing beam splitter is a critical beam-splitting element in space optics. A high-splitting-ratio unpolarizing beam splitter was designed and fabricated in this study to achieve a stable output of a 780 nm band laser with a large splitting ratio. First， the double-sided unpolarizing film system was simulated and designed using TFCalc. Next， unpolarizing beam splitter samples were fabricated using ion beam-assisted deposition. Subsequently， the samples were measured and characterized using transmission electron microscopy and spectrophotometry， and the actual film structure and transmission spectrum were obtained. The spectral results showed that the transmittance of the beam splitter was close to 98%， and the transmittance deviation |Ts -Tp| was lower than 0.3%. Finally， the actual performance of the beam splitter was experimentally evaluated. The experimental results show that the transmittance of the beam splitter is close to 98% in the target wavelength band， and the transmittance deviation |Ts -Tp| is lower than 0.2% when the polarization direction changed in a broad range. The transmittance fluctuation is lower than 0.12% at 772-792 nm. The 10 h long-term results indicated that when the average time is 100 s， the Allan variances of the split ratio and transmittance are 1.4×10-3 and 4.12×10-5， respectively. The proposed unpolarizing beam splitter exhibits good depolarization performance and can be directly applied to precision measurement fields， such as optical test metrology and quantum-sensing detection.关键词：nonpolarizing;beam splitter;film system design;high splitting ratio
- 64
- |
- 12
- |
- 0
发布时间：2023-11-16 - 摘要：In recent years， digital micromirror device （DMD）， as a flexible， programmable， and independently addressable spatial light modulation device， has been widely used in maskless lithography， beam shaping， holographic imaging， confocal measurement， and other fields. The flexible modulation mode also gives DMD unique advantages in spectral imaging. It can flexibly modulate the imaging field of view， replacing the traditional mechanical mask and scanning structure. In this study， DMD's research progress and application in spectral imaging in recent years was reviewed. The basic structure and working principle of the optical system of the aperture-coded and push-broom spectral imaging system based on DMD were discussed in detail and the development of a spectral imaging system based on DMD from Hadamard transform spectral imaging to push-broom spectral imaging was summarized. The relevant research work to overcome the DMD's diffraction owing to the micro-mirror and inclination of the image plane was introduced. Finally， The unique advantages of spectral imaging technology based on DMD was summarized and the future development direction and application prospect of spectral imaging technology based on DMD was discussed .关键词：digital micromirror device;spectral imaging;coded aperture;Hadamard transform;push-broom
- 33
- |
- 16
- |
- 0
发布时间：2023-11-16
Modern Applied Optics
- 摘要：A real-time error measurement system was developed to measure the motion error of a two-dimensional linear module. The system consisted of a four-degree-of-freedom motion error measurement module （for measuring the horizontal straightness， vertical straightness， pitch angle， and yaw angle errors）， roll angle error measurement module （for measuring the roll angle error）， and linear grating ruler （for measuring the axial positioning error）； thus， it measured six degrees of freedom motion errors in a single axis. The principle of the homogeneous transformation matrix （HTM）， was used to construct a spatial error model for the two-dimensional module to represent the actual spatial positions of functional points. The calibration experiments of the measurement system were performed， and the experimental data were processed based on the Abbe-Bryan principle to complete the comparison experiment. Finally， the positioning， straightness， and measurement accuracies of angular errors reach ±1.2 μm， ±1.3 μm， and ±1''， respectively. The measurement error of the diagonal position in the XZ plane of the two-dimensional linear module was analyzed based on the proposed spatial error model. The results indicate that the measurement error of the diagonal position in the XZ plane decreases from 68 μm to 13 μm after using the two-dimensional linear module spatial error model. In addition， because the motion errors of the two-dimensional linear module at different positions changes under loading， a comparative experiment was conducted by adding a standard weight of 2 kg to the Z-axis slider to verify the measurement system could capture real-time spatial error variations of the linear module. The results show that the measurement error of the diagonal position in the XZ plane decreases from 56 μm to 14 μm after using the two-dimensional linear module spatial error model.关键词：error measurement system;two-dimensional linear module;spatial error model;real-time measurement
- 28
- |
- 15
- |
- 0
发布时间：2023-11-16 - 摘要：The geometric error of the rotary axis significantly impacts the machining accuracy of the five-axis machine tool. However， its identification is challenging due to the multitude of error terms and high coupling. This paper introduces a novel machining test and on-machine measurement method designed to identify the six position-dependent geometric errors （PDGE） of the rotary axis. Initially， a misaligned tower-shaped artifact， consisting of three layers of misaligned superimposed rectangular blocks， was designed and processed. Subsequently， measurement points were strategically arranged on the bottom and sides of the artifact at different levels， and on-machine measurements were executed. Using the volumetric error model， the identification principle and analytical solution for each error were derived， and uncertainty analysis was conducted through Monte Carlo simulation. Comparing with the ball-bar error identification method， the maximum deviation of the linearity errors EXC， EYC and EZC are 2.7 μm， -1.7 μm and -1.3 μm， respectively. Meanwhile， the maximum deviation of the angle errorsthe maximum deviation of the angle errors EAC， EBC and ECC are 1.3"， -0.6"， and -2.1"， respectively， with an overall agreement degree of 95.4% for the PDGE. Through the machining test and on-machine measurement， the identification principles and analytical solutions for each error are straightforward， allowing for the identification of the six PDGE of the rotary axis under actual working conditions.关键词：five-axis machine tool;rotary axis;geometric error;error identification;machining test;on-machine measurement
- 18
- |
- 11
- |
- 0
发布时间：2023-11-16
Micro/Nano Technology and Fine Mechanics
- 摘要：In order to meet the battery replacement demand of electric vehicles， a 6D pose estimation method of battery package locking mechanism based on point cloud segmentation is proposed to solve the positioning problem of locking mechanism during battery package docking in battery swapping station. This method uses YOLOv5 network to segment the point cloud of locking mechanism from the scene， and uses voxel filtering and moving least square fitting to filter and smooth the point cloud. The point cloud labels are predicted by the point cloud segmentation network， and the global semantic feature is added to the Fast Point Feature Histograms （FPFH） feature to make up for the defect that the FPFH has only the local feature of the point cloud. According to this feature， the Random Sample Consensus （RANSAC） rigid point cloud registration is leveraged， and the 6D pose of the locking mechanism point cloud is estimated. Finally， the Iterative Closest Point（ICP） algorithm is used to correct the pose estimation results. The experimental results show that the 6D pose estimation algorithm of locking mechanism based on point cloud segmentation has high accuracy， and can overcome the mismatching caused by environmental noise， and accurately obtain the position and attitude of locking mechanism. The angle error of position and attitude estimation can reach 1.90°， the displacement error can reach 1.4 mm， and the RMSE can reach 1.5 mm， which provides an effective solution for battery docking positioning in battery swapping station.关键词：Battery Swapping robot;battery pack;pose estimation;point cloud registration;point cloud segmentation
- 19
- |
- 5
- |
- 0
发布时间：2023-11-16 - 摘要：As only one or a few training samples are used for few-shot classification tasks， the features extracted via a prototypical network cannot guarantee much discriminative power. Accordingly， this paper proposes an intra-inter channel attention few-shot classification （ICAFSC） method. This method uses an intra–inter channel attention module （ICAM） to calculate channel weights based on an intra-inter distance metric. The module is integrated into the prototypical network to make the embedded features more discriminative. To overcome the problems of overfitting or underfitting when directly learning the ICAM in the few-shot classification's meta-training stage， ICAFSC adds a pre-training stage before the meta-training of the prototypical network. We design adequate classification tasks with a large number of labeled samples to learn optimal parameters of the ICAM in the pre-training stage. Subsequently， in the meta-training and meta-testing stages of the prototypical network， ICAFSC first freezes the parameters of the ICAM to guarantee a stable channel attention relationship. It then achieves few-shot classification experience learning and transfer via meta-training and meta-testing. We conduct 1-shot and 5-shot few-shot classification experiments on the MiniImagenet dataset. The experimental results indicate that， compared to the prototypical network， the proposed ICAFSC method shows improvements of 1.93% and 1.15% for the 1-shot and 5-shot scenarios， respectively.关键词：deep learning;few-shot classification;meta-learning;prototypical network;channel attention
- 157
- |
- 10
- |
- 0
发布时间：2023-11-16 - 摘要：The statistics of background information in hyperspectral target detection are often interfered by target information， and the presence of a large number of mixed pixels in hyperspectral images will further deepen this interference. In this study， we proposed a target detection algorithm using spectral unmixing to accurately calculate background information and significantly reduce the interference of target pixels on background statistical information. First， we obtained the abundance coefficient corresponding to the target end member by spectral unmixing and target similarity judgment. We combined it with the spectral angle coefficient to generate a reasonable background weighting coefficient for weighted constrained energy minimization （CEM） target detection， effectively improving the statistical accuracy of background information of mixed pixels. Second， we generated a preliminary result of target detection by utilizing the abundance coefficient corresponding to the target end member and spectral angle coefficient and fused with the weighted CEM target detection result to optimize further， effectively improving the robustness of the algorithm and target detection accuracy. Experimental results showed that the algorithm proposed in this study has good target detection performance for simulated or real hyperspectral images. The algorithm has strong robustness and effectively improves target detection accuracy. Compared with the traditional CEM algorithm， weighted CEM algorithm based on spectral angle， and normalized abundance coefficient as the target result the AUC of this study was promoted by an average of 0.071 2， 0.031 2， and 0.015 0， respectively. The proposed algorithm has strong practicability in hyperspectral applications.关键词：hyperspectral image;target detection;hyperspectral unmixing;abundance;spectral angle
- 22
- |
- 9
- |
- 0
发布时间：2023-11-16 - 摘要：Light-field （LF） imaging can capture spatial and angular information of light rays in a scene. Compared to traditional 2D/3D images， LF images provide a more comprehensive description of the scene. To address the problem of low angular resolution in LF images caused by hardware constraints， a LF angular super-resolution reconstruction method based on the fusion of 3D epipolar plane images （EPIs） is proposed. First， to make full use of the parallax information of the input images and improve the accuracy of depth estimation， the input images are arranged in varying parallax directions， and their features are extracted individually. Then， initial synthetic LF images are generated by transforming the input images to match the new viewpoint location using the corresponding depth maps. Finally， to ensure that the reconstructed LF image retains better detail information and geometric consistency， the LF is reconstructed horizontally and vertically via the horizontal and vertical 3D EPI fusion reconstruction branches， respectively. These two reconstruction results are then fused to produce the final high-angular-resolution LF image. Experimental results show that， compared to existing methods， the proposed method achieves an improved reconstruction quality across both synthetic and real-world LF image datasets， and the maximum increase in the peak signal-to-noise ratio reaches 1.99%. Thus， the proposed method can effectively improve the quality of the reconstructed LF.关键词：light field;super-resolution reconstruction;3D epipolar plane image;convolution neural network
- 14
- |
- 9
- |
- 0
发布时间：2023-11-16 -
Mandibular fracture detection with 3M-YOLOv5 network based on enhanced feature extraction capability
摘要：For artificial intelligence assistance in the detection of fracture sites， the fracture sites are usually accompanied by bleeding and other symptoms. Further， CT images taken in different positions have large differences， the size of fracture sites varies， and the bleeding sites and surrounding tissues may interfere with the detection of fracture sites， leading to insufficient feature extraction and the problem of low detection accuracy. Therefore， the 3M-YOLOv5 network is designed to detect mandibular fracture sites. First， the dense module is used in the feature extraction network to improve the feature extraction capability of the network by using the dense connection property. The local and global attention module （lgaM） is used to extract the global information of CT images. Second， a lightweight multiscale dense block （lmdM） is designed to extract the multiscale features of the fracture sites with fewer parameters. Third， a cross-dimension bidirectional feature fusion module （cdbfM） is designed in the feature enhancement network to make the height， width， and channel of the feature maps interact with each other， and trainable weights are introduced to balance the fusion importance of the feature maps with different scales. Finally， to verify the effectiveness of the proposed network， ablation and comparison experiments are conducted on a self-built dataset. The results show that when the confidence threshold is 0.5， the mAP value， F1 value， recall rate， and precision rate of the proposed network are 99.17%， 99.06%， 98.81%， and 99.32%， respectively. The proposed CT image detection network for mandibular fracture can better detect the fracture sites in the image than existing methods， which is a good reference for doctors to make a corresponding treatment plan based on the detection results.关键词：target detection;mandibular fracture;YOLOv5;cross dimension attention;densely connected neural network- 17
- |
- 12
- |
- 0
发布时间：2023-11-16 - 摘要：Traditional face recognition methods have poor recognition performance； deep learning-based methods face difficulty recognizing under unrestricted conditions， face features are weakly differentiated， and recognition accuracy is easily affected by pose and expression. To address this， a twin neural network model structure that introduces a Convolutional Block Attention Module （CBAM） hybrid attention mechanism is proposed. First， the algorithm structure based on the basic framework of the Siamese neural network was improved and the improved VGG11 into the framework is introduced. The BN model for feature extraction is used， which introduces batch normalization （BN） technology based on the VGG11 structure. Second， a feature extraction network incorporating a CBAM mixed attention mechanism was introduced based on the original model structure. Finally， in response to the lack of facial recognition training for Asians， the CASIA-FaceV5 dataset was employed， which is more aligned with Asian facial features， for recognition training. The experimental results show that the algorithm’s accuracy reaches 96.67% in face recognition， and the accuracy on CAS-PEAL-R1 face dataset is 6.05% and 6.7% higher than that of SRGES and VGG11+siamese algorithms， respectively. The algorithm in this study can better verify facial recognition under multiple factors， has good robustness， and greater application value.关键词：face recognition;Siamese neural network;deep learning;attention mechanism;robustness
- 19
- |
- 9
- |
- 0
发布时间：2023-11-16 - 摘要：To suppress irrelevant semantics and cross semantic gaps in object extraction using an encoder-decoder network structure， thereby achieving higher accuracy. U-Net is used as the backbone network for feature extraction. To reduce semantic differences between shallow and deep features， a multi-scale semantic pooling module （CSP， Channel-Spatial-Pyramid） integrates attention perception and replaces skip links in early layers. The CSP module emphasizes more meaningful semantic information from two levels corresponding to space and channel， extracts features at different scales through parallel branches of four different pooling cores， and aggregates all branch results to splice with the features of later layers. The experimental results show that the Dice index of CSP-Net in color fundus image disc segmentation reaches 99.6%， whereas that of cup segmentation reaches 92.1%. Both results represent improvements over existing algorithms. CSP-Net exhibits a high effectiveness and anti-interference ability for extracting small targets in fundus images， making it appropriate for clinical reference in glaucoma screening and diagnosis.关键词：Multi scale semantics;Attention perception;target extraction;U-net
- 12
- |
- 8
- |
- 0
发布时间：2023-11-16 - 摘要：As light propagation in water is subject to absorption and scattering effects， acquiring underwater images using conventional intensity cameras can result in low brightness of imaging results， blurred images， and loss of details. In this study， a deep fusion network was applied to underwater polarimetric images； the underwater polarimetric images were fused with light-intensity images using deep learning. First， the underwater active polarization imaging model was analyzed， an experimental setup was built to obtain underwater polarization images to construct a training dataset， and appropriate transformations were performed to expand the dataset. Second， an end-to-end learning framework was constructed based on unsupervised learning and guided by attention mechanism for fusing polarimetric and light intensity images and the loss function and weight parameters were elaborated. Finally， the produced dataset was used to train the network under different loss weight parameters and the fused images were evaluated based on different image evaluation metrics. The experimental results show that the fused underwater images are more detailed， with 24.48% higher information entropy and 139% higher standard deviation than light-intensity images. Unlike other traditional fusion algorithms， the method does not require manual weight parameter adjustment， has faster operation speed， strong robustness， and self-adaptability， which is important for ocean detection and underwater target recognition.关键词：polarization imaging;underwater image enhancement;image fusion;deep learning
- 12
- |
- 13
- |
- 0
发布时间：2023-11-16 - 摘要：Herein， a feature extraction method based on fractional differentiation is proposed for the feature extraction and classification of hyperspectral images. Two-dimensional （2D） fractional differential masks are designed to extract the pixel spatial fractional differential （SpaFD） feature of hyperspectral images， and a spectral–spatial joint criterion is proposed to select the differential mask order. To entirely utilize the spatial and spectral features of hyperspectral images， the SpaFD feature is fused with the original feature via a direct connection to obtain a mixed feature （SpaFD-Spe-Spa）. The effectiveness of the SpaFD-Spe-Spa feature is verified on a 3D convolutional neural network （3DCNN）， 3DCNN after pixel spectrum dimensionality reduction using principal component analysis （3DCNNPCA）， and hybrid spectral network （HybridSN）. In the experiment， masks with sizes of 3×3， 5×5， and 7×7 are used to perform feature extraction. Experiments on four real hyperspectral image datasets reveal that the extracted SpaFD and SpaFD-Spe-Spa features are effective in hyperspectral image classification， and the SpaFD-Spe-Spa feature significantly improves classification accuracy. When compared with the original features in the Indian Pines， Botswana， Pavia University， and Salinas datasets， the classification accuracy of the SpaFD feature is improved by 3.87%， 1.42%， 2.41%， and 2.87%， respectively， whereas that of the SpaFD-Spe-Spa feature is improved by 3.90%， 5.62%， 3.35%， and 5.18%， respectively， under optimal conditions.关键词：hyperspectral images classification;fractional differentiation;feature extraction;convolutional neural network
- 16
- |
- 6
- |
- 0
发布时间：2023-11-16
Information Sciences
- Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China Postal code：130033
- Tel：0431-86176855 Email：gxjmgc@ciomp.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0