Binocular stereo vision is a vital research topic in computer vision and has been widely used in robot navigation, 3d reconstruction and other fields. Stereo matching is the most critical part of binocular stereo vision. In view of the low computational efficiency of local stereo matching methods, non-local methods based on tree structure have attracted much attention of researchers in recent years. In this paper, we propose a new efficient non-local spatial tree filter (NSTF) to aggregate the matching cost. Firstly, in addition to spatial affinity, the internal color similarity is taken as the similarity measure between adjacent pixels. Then, the propagation of cost is carried out recursively in the form of ternary tree. The whole filtering process can be divided into eight different directions. Quantitative experiments on Middlebury benchmark show that NSTF can effectively improve the accuracy of the algorithm, and has better edge-preserving ability than other tree-based non-local methods, especially in weak texture and high texture regions.
{"title":"An efficient stereo matching method based on non-local spatial tree filter","authors":"He Zhong, Yahu Zhu, Deqi Ming","doi":"10.1145/3548608.3559180","DOIUrl":"https://doi.org/10.1145/3548608.3559180","url":null,"abstract":"Binocular stereo vision is a vital research topic in computer vision and has been widely used in robot navigation, 3d reconstruction and other fields. Stereo matching is the most critical part of binocular stereo vision. In view of the low computational efficiency of local stereo matching methods, non-local methods based on tree structure have attracted much attention of researchers in recent years. In this paper, we propose a new efficient non-local spatial tree filter (NSTF) to aggregate the matching cost. Firstly, in addition to spatial affinity, the internal color similarity is taken as the similarity measure between adjacent pixels. Then, the propagation of cost is carried out recursively in the form of ternary tree. The whole filtering process can be divided into eight different directions. Quantitative experiments on Middlebury benchmark show that NSTF can effectively improve the accuracy of the algorithm, and has better edge-preserving ability than other tree-based non-local methods, especially in weak texture and high texture regions.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123997568","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
According to the survey data about the waters of the Jiaozhou Bay from 1986 to 1990, through the changing trend of spatial sedimentation of Hg and the mechanism of the spatial sedimentation based on variation range of Hg content in both surface and bottom waters and the horizontal distribution trend of Hg content in both surface and bottom waters were adequately demonstrated. Therefore, With our research on the trends of both the surface and bottom horizontal distribution from 1986 to 1990, we found that from the start point, a position where the source of Hg was close to the surface horizontal distribution, to the end point, a position where the source of Hg was far from the surface horizontal distribution, the content of Hg decreased in gradient order. While the trend of the horizontal distribution in bottom waters, from the beginning to the end, had two changes: (1) gradient descent and (2) gradient rise. With all these efforts, the author proposed Dongfang Yang rules of the changes of Hg content in surface waters: no matter where it came from and how much it was, as long as the place stayed away from the source of Hg, the Hg content there in surface waters continued decreasing and to be deposited. When there was only one channel, the Hg content in surface waters conformed to the Dongfang Yang rules of the changes of Hg content in surface waters. When there were two ways of Hg transportation or more, regarding the major source as the start point, the Dongfang Yang rules of the changes of Hg content in surface waters could work here. At that time, the Hg content in the bottom waters didn't accumulated. Thereby, from the start point to the end point of the horizontal distribution of Hg, the Hg content experienced a gradient decease, showing the consistency of the trends of horizontal distribution of in both surface and bottom waters. As the Hg in the surface waters further decreased and was deposited, the Hg in the bottom waters began accumulating in a large amount. Thereby, from the start point to the end point of horizontal distribution of Hg in the bottom waters, the trend of the horizontal distribution of Hg content increased in gradient order, indicating that the trend of the horizontal distribution in the surface waters was opposite against that in the bottom waters.
{"title":"The horizontal distribution trend and changing rules of Hg content: In the application of the data fusion","authors":"Dongfang Yang, Junchao Wei","doi":"10.1145/3548608.3559329","DOIUrl":"https://doi.org/10.1145/3548608.3559329","url":null,"abstract":"According to the survey data about the waters of the Jiaozhou Bay from 1986 to 1990, through the changing trend of spatial sedimentation of Hg and the mechanism of the spatial sedimentation based on variation range of Hg content in both surface and bottom waters and the horizontal distribution trend of Hg content in both surface and bottom waters were adequately demonstrated. Therefore, With our research on the trends of both the surface and bottom horizontal distribution from 1986 to 1990, we found that from the start point, a position where the source of Hg was close to the surface horizontal distribution, to the end point, a position where the source of Hg was far from the surface horizontal distribution, the content of Hg decreased in gradient order. While the trend of the horizontal distribution in bottom waters, from the beginning to the end, had two changes: (1) gradient descent and (2) gradient rise. With all these efforts, the author proposed Dongfang Yang rules of the changes of Hg content in surface waters: no matter where it came from and how much it was, as long as the place stayed away from the source of Hg, the Hg content there in surface waters continued decreasing and to be deposited. When there was only one channel, the Hg content in surface waters conformed to the Dongfang Yang rules of the changes of Hg content in surface waters. When there were two ways of Hg transportation or more, regarding the major source as the start point, the Dongfang Yang rules of the changes of Hg content in surface waters could work here. At that time, the Hg content in the bottom waters didn't accumulated. Thereby, from the start point to the end point of the horizontal distribution of Hg, the Hg content experienced a gradient decease, showing the consistency of the trends of horizontal distribution of in both surface and bottom waters. As the Hg in the surface waters further decreased and was deposited, the Hg in the bottom waters began accumulating in a large amount. Thereby, from the start point to the end point of horizontal distribution of Hg in the bottom waters, the trend of the horizontal distribution of Hg content increased in gradient order, indicating that the trend of the horizontal distribution in the surface waters was opposite against that in the bottom waters.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124889250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zheyu Zhang, QinLi Liu, Jiao Li, Xinyao Gong, Dongli Liu
Precision fishing and real-time monitoring are growing trends toward "intelligent" fishery. However, the fishery is still in a relatively backward artificial aquaculture due to the complex underwater environment and the intricate movement posture of fish. To realize the precision aquaculture of fishery, this paper took golden crucian carp for example selecting 10 golden crucian carp as experimental objects and formed 640 golden crucian carp instance segmentation data set through labelme manual labelling method. This paper proposes a method to solve fish segmentation and distinguish individuals. The solution optimises the cascade mask RCNN model, so that the model could effectively perform case segmentation of gold crucian carp. By comparing the predictive models and multiple case segmentation models, the former performs better. Among them, the mAP of Bbox of Cascade Mask RCNN reached 0.916, and the mAP of segmentation also reached 0.917, which can effectively complete the task of individual differentiation and body size estimation for Gold crucian carp. This study provides a data set and reference for body size estimation in fish farming.
{"title":"Instance Segmentation of Golden crucian carp based on Cascade Mask RCNN","authors":"Zheyu Zhang, QinLi Liu, Jiao Li, Xinyao Gong, Dongli Liu","doi":"10.1145/3548608.3559183","DOIUrl":"https://doi.org/10.1145/3548608.3559183","url":null,"abstract":"Precision fishing and real-time monitoring are growing trends toward \"intelligent\" fishery. However, the fishery is still in a relatively backward artificial aquaculture due to the complex underwater environment and the intricate movement posture of fish. To realize the precision aquaculture of fishery, this paper took golden crucian carp for example selecting 10 golden crucian carp as experimental objects and formed 640 golden crucian carp instance segmentation data set through labelme manual labelling method. This paper proposes a method to solve fish segmentation and distinguish individuals. The solution optimises the cascade mask RCNN model, so that the model could effectively perform case segmentation of gold crucian carp. By comparing the predictive models and multiple case segmentation models, the former performs better. Among them, the mAP of Bbox of Cascade Mask RCNN reached 0.916, and the mAP of segmentation also reached 0.917, which can effectively complete the task of individual differentiation and body size estimation for Gold crucian carp. This study provides a data set and reference for body size estimation in fish farming.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129742848","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
An attitude adjustment device for placing the neutron monochromator which is the important component of the stress spectrometer is designed. It has the characteristics of large mechanical transmission ratio, diversified transmission modes and large load. Using the open-loop control system to drive the stepper motor has the problems of step loss and large mechanical transmission gap, and the positional positioning accuracy is difficult to guarantee; due to the influence of the mechanical transmission gap, friction and load inertia ,using the closed-loop control system based on the PI control method will cause nonlinear natural vibration of the system; a composite closed-loop control method based on stepper motor is proposed, which improves the stability of the system, overcomes the mechanical transmission error, and improves the positioning accuracy.
{"title":"An Attitude Adjustment Device for the Neutron Monochromator","authors":"Kangkang An, G. Ma, Bin Liu","doi":"10.1145/3548608.3561136","DOIUrl":"https://doi.org/10.1145/3548608.3561136","url":null,"abstract":"An attitude adjustment device for placing the neutron monochromator which is the important component of the stress spectrometer is designed. It has the characteristics of large mechanical transmission ratio, diversified transmission modes and large load. Using the open-loop control system to drive the stepper motor has the problems of step loss and large mechanical transmission gap, and the positional positioning accuracy is difficult to guarantee; due to the influence of the mechanical transmission gap, friction and load inertia ,using the closed-loop control system based on the PI control method will cause nonlinear natural vibration of the system; a composite closed-loop control method based on stepper motor is proposed, which improves the stability of the system, overcomes the mechanical transmission error, and improves the positioning accuracy.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129905801","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In order to help patients with lower extremity rehabilitation training, this project aims to design a lower extremity-assisted and adjustable exoskeleton robot. First, the degree of freedom of each joint of the exoskeleton was determined according to the characteristics of human anatomy and bone connection. and the structural design of the normal size of the mechanism was determined by the average size of the bones. Secondly, the angle equation determined by the leg joint coordinates is established, and the angle parameters of different time periods in a gait cycle are obtained by observing the changes of the joint coordinates. The coordinate system of each joint is established by the DH parameter method. It provides a theoretical basis for animation simulation, structural design, kinematics analysis and control. Finally, the kinematics analysis and simulation are carried out in MATLAB; the overall structure and simulation animation of the robot are established by SOLIDWORKS. The PID method is used to construct the power-assisted control frame diagram of the robot. The results showed that the exoskeleton robot responds well to various kinematics and other properties of the lower limbs.
{"title":"Design and simulation of a lower limb exoskeleton rehabilitation robot","authors":"Wenkai Qiao, Yapeng Tong, Zilu Zhang","doi":"10.1145/3548608.3559176","DOIUrl":"https://doi.org/10.1145/3548608.3559176","url":null,"abstract":"In order to help patients with lower extremity rehabilitation training, this project aims to design a lower extremity-assisted and adjustable exoskeleton robot. First, the degree of freedom of each joint of the exoskeleton was determined according to the characteristics of human anatomy and bone connection. and the structural design of the normal size of the mechanism was determined by the average size of the bones. Secondly, the angle equation determined by the leg joint coordinates is established, and the angle parameters of different time periods in a gait cycle are obtained by observing the changes of the joint coordinates. The coordinate system of each joint is established by the DH parameter method. It provides a theoretical basis for animation simulation, structural design, kinematics analysis and control. Finally, the kinematics analysis and simulation are carried out in MATLAB; the overall structure and simulation animation of the robot are established by SOLIDWORKS. The PID method is used to construct the power-assisted control frame diagram of the robot. The results showed that the exoskeleton robot responds well to various kinematics and other properties of the lower limbs.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"2493 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127478459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zhiyuan Cheng, Zhiguo Li, Zhou Ji, Peipei Yan, Ai-li Xia
The laser field imaging telescope emits laser coherent field to detect space targets. When laser passes through the atmospheric transmission path, the laser is inevitably affected by the phase jitter of atmosphere. Atmospheric phase jitter varies randomly and it is difficult to be quantitatively simulated. In order to quantitatively research the influence of atmospheric phase jitter on imaging quality, a quantitative simulation method of atmospheric phase jitter based on the change of beam frequency difference was proposed. A set of atmospheric turbulence phase jitter quantitative numerical simulation experimental platform was constructed. The effect of different phase jitter on the imaging quality of laser field imaging was quantitatively studied. The results show that the larger the atmospheric phase jitter is, the more degraded the image quality is. We suggest that the atmospheric phase jitter should be suppressed in the signal reconstruction. The quantitative simulation method of atmospheric phase jitter provides an effective technique for studying the image quality of light field imaging.
{"title":"Quantitative atmospheric phase jitter simulating research of optic field imaging","authors":"Zhiyuan Cheng, Zhiguo Li, Zhou Ji, Peipei Yan, Ai-li Xia","doi":"10.1145/3548608.3559198","DOIUrl":"https://doi.org/10.1145/3548608.3559198","url":null,"abstract":"The laser field imaging telescope emits laser coherent field to detect space targets. When laser passes through the atmospheric transmission path, the laser is inevitably affected by the phase jitter of atmosphere. Atmospheric phase jitter varies randomly and it is difficult to be quantitatively simulated. In order to quantitatively research the influence of atmospheric phase jitter on imaging quality, a quantitative simulation method of atmospheric phase jitter based on the change of beam frequency difference was proposed. A set of atmospheric turbulence phase jitter quantitative numerical simulation experimental platform was constructed. The effect of different phase jitter on the imaging quality of laser field imaging was quantitatively studied. The results show that the larger the atmospheric phase jitter is, the more degraded the image quality is. We suggest that the atmospheric phase jitter should be suppressed in the signal reconstruction. The quantitative simulation method of atmospheric phase jitter provides an effective technique for studying the image quality of light field imaging.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130719206","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper studies a collaborative robot obstacle avoidance method based on digital twin technology, by establishing a digital twin of the robot, collecting and transmitting the twin data, so that the virtual model and the robot entity can map each other to realize the virtual reality and real-time monitoring of the robot operation state. A robot obstacle avoidance path planning method based on the improved artificial potential field method is proposed. Find the closest point of each linkage of the robot to the obstacle and add the reverse control force based on the potential field function to achieve obstacle avoidance of the robot. The simulation debugging is carried out on the Unity 3D software platform, and the results show that the robot can achieve collision-free motion in the process of reaching the target point, which verifies the feasibility of the method.
{"title":"Research on Robot Obstacle Avoidance Method Based on Digital Twin","authors":"Lei Wang, Yongfei Xiao","doi":"10.1145/3548608.3559179","DOIUrl":"https://doi.org/10.1145/3548608.3559179","url":null,"abstract":"This paper studies a collaborative robot obstacle avoidance method based on digital twin technology, by establishing a digital twin of the robot, collecting and transmitting the twin data, so that the virtual model and the robot entity can map each other to realize the virtual reality and real-time monitoring of the robot operation state. A robot obstacle avoidance path planning method based on the improved artificial potential field method is proposed. Find the closest point of each linkage of the robot to the obstacle and add the reverse control force based on the potential field function to achieve obstacle avoidance of the robot. The simulation debugging is carried out on the Unity 3D software platform, and the results show that the robot can achieve collision-free motion in the process of reaching the target point, which verifies the feasibility of the method.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130239315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the agriculture industry, scholars are increasingly paying attention to the advantages of hybrid robots and applying them in agricultural production. Agricultural products such as tree trunks or fruit stems is formed by growth. When the hybrid robot is working, It is easier to obtain the ideal end pose of the hybrid robot by describing the pose relationship of the robot end relative to the work object than the robot end relative to the world coordinate system. In order to realize the continuous motion control of the hybrid robot with a specified relative attitude relative to a certain plane in space, a method to realized planar motion based on 2SPU+U+RRR type five-degree-of-freedom hybrid robot position and attitude control is proposed. First, the trajectory of the task is discretized, according to the required posture, and the position and attitude matrix of each position on the trajectory is derived. Then through the inverse solution of the kinematic model, the five-degree-of-freedom hybrid robot is completely decoupled. The pose of the robot end is calculated to realize the plane motion control of the robot with a specified pose in Cartesian space according to the kinematic model of the hybrid robot based on the mechanism equivalent principle. Finally, the pose control method of the robot is tested. The experimental results show that the control method can realize the accurate planar motion control of the specified pose in Cartesian space during the repeated planar motion of the 2SPU+U+RRR type 5-DOF hybrid robot. This paper provides a five-degree-of-freedom hybrid robot theoretical model for the position and attitude synchronous motion control of arbitrary plane in Cartesian space.
{"title":"Research on position and attitude control method of planar motion of 5-DOF hybrid robot","authors":"Taiyu Wu, Jin Gu, Yawei Zhang, Bin Zhang","doi":"10.1145/3548608.3559184","DOIUrl":"https://doi.org/10.1145/3548608.3559184","url":null,"abstract":"In the agriculture industry, scholars are increasingly paying attention to the advantages of hybrid robots and applying them in agricultural production. Agricultural products such as tree trunks or fruit stems is formed by growth. When the hybrid robot is working, It is easier to obtain the ideal end pose of the hybrid robot by describing the pose relationship of the robot end relative to the work object than the robot end relative to the world coordinate system. In order to realize the continuous motion control of the hybrid robot with a specified relative attitude relative to a certain plane in space, a method to realized planar motion based on 2SPU+U+RRR type five-degree-of-freedom hybrid robot position and attitude control is proposed. First, the trajectory of the task is discretized, according to the required posture, and the position and attitude matrix of each position on the trajectory is derived. Then through the inverse solution of the kinematic model, the five-degree-of-freedom hybrid robot is completely decoupled. The pose of the robot end is calculated to realize the plane motion control of the robot with a specified pose in Cartesian space according to the kinematic model of the hybrid robot based on the mechanism equivalent principle. Finally, the pose control method of the robot is tested. The experimental results show that the control method can realize the accurate planar motion control of the specified pose in Cartesian space during the repeated planar motion of the 2SPU+U+RRR type 5-DOF hybrid robot. This paper provides a five-degree-of-freedom hybrid robot theoretical model for the position and attitude synchronous motion control of arbitrary plane in Cartesian space.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127875964","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Compared with traditional electrical communication, optical fiber communication has many advantages in signal transmission. It has the characteristics of wide transmission frequency bandwidth, low loss and strong anti-interference ability. Optical fiber communication technology has a wide application prospect in civil aircraft. This paper introduces the application of optical fiber communication technology in civil aircraft. At present, optical fiber communication is mainly used in civil aircraft weather radar, Enhanced Flight Vision system, inertial navigation system, integrated modular avionics system and so on. This paper also looks forward to the large-scale use of optical fiber communication technology in civil aircraft in the future. It can greatly improve the aircraft communication efficiency, reduce the weight of the aircraft and improve the available space of the aircraft.
{"title":"The Application and Expectation of Optical Fiber Communication Used on Civil Aircraft","authors":"Pei-yu Wei","doi":"10.1145/3548608.3559312","DOIUrl":"https://doi.org/10.1145/3548608.3559312","url":null,"abstract":"Compared with traditional electrical communication, optical fiber communication has many advantages in signal transmission. It has the characteristics of wide transmission frequency bandwidth, low loss and strong anti-interference ability. Optical fiber communication technology has a wide application prospect in civil aircraft. This paper introduces the application of optical fiber communication technology in civil aircraft. At present, optical fiber communication is mainly used in civil aircraft weather radar, Enhanced Flight Vision system, inertial navigation system, integrated modular avionics system and so on. This paper also looks forward to the large-scale use of optical fiber communication technology in civil aircraft in the future. It can greatly improve the aircraft communication efficiency, reduce the weight of the aircraft and improve the available space of the aircraft.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127901484","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In order to solve the problem that embedding attention mechanism module into lightweight convolutional neural network will increase the number of parameters and FLOPs with less accuracy improvement, a Teeth Attention Module (TAM) based on lightweight network was proposed. The feature map is enhanced from two aspects of channel attention and spatial attention. In the channel attention module, local channel interaction is carried out from height and width without dimensionality reduction through one-dimensional convolution, and it is divided into MP-ECAM and AP-ECAM according to pooling mode. In the spatial attention module, the average value and maximum value of each pixel point in the feature map are calculated according to the dimensions to construct ESAM. Finally, TAM is constructed in a channel-space-channel manner. TAM was embedded into the lightweight networks MobileNetV2, ShuffleNetV2 and EfficientNetV1 respectively, and the experiment was carried out on CIFAR-10 image classification datasets. Compared with Squeeze and Excitation Module (SE), Convolutional Block Attention Module (CBAM) and Efficient Channel Attention Module (ECA), this module can achieve higher accuracy with less increase of the number of parameters and FLOPs.
{"title":"Teeth attention mechanism image classification based on lightweight network","authors":"Zejun Zhang, Yulong He, Huabin He, Zhiming Cai","doi":"10.1145/3548608.3559211","DOIUrl":"https://doi.org/10.1145/3548608.3559211","url":null,"abstract":"In order to solve the problem that embedding attention mechanism module into lightweight convolutional neural network will increase the number of parameters and FLOPs with less accuracy improvement, a Teeth Attention Module (TAM) based on lightweight network was proposed. The feature map is enhanced from two aspects of channel attention and spatial attention. In the channel attention module, local channel interaction is carried out from height and width without dimensionality reduction through one-dimensional convolution, and it is divided into MP-ECAM and AP-ECAM according to pooling mode. In the spatial attention module, the average value and maximum value of each pixel point in the feature map are calculated according to the dimensions to construct ESAM. Finally, TAM is constructed in a channel-space-channel manner. TAM was embedded into the lightweight networks MobileNetV2, ShuffleNetV2 and EfficientNetV1 respectively, and the experiment was carried out on CIFAR-10 image classification datasets. Compared with Squeeze and Excitation Module (SE), Convolutional Block Attention Module (CBAM) and Efficient Channel Attention Module (ECA), this module can achieve higher accuracy with less increase of the number of parameters and FLOPs.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126846552","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：