Pub Date : 2022-08-07DOI: 10.1109/ICMA54519.2022.9856048
Shuizhong Zou, Yuan Huang, Ziang Wang
In the master-slave operation stage of robot minimally invasive surgery, doctors need to frequently switch the control objects to adjust the position and posture of the laparoscope to obtain a better surgical field of vision, which will distract doctors' attention and lead to a decrease in the quality of laparoscopic surgery. Therefore, a laparoscopic automatic following motion planning is proposed to track the motion of surgical instruments quickly, smoothly and safely. Firstly, the position and posture adjustment conditions and adjusted visual field requirements of the distal reference point of the laparoscope are analyzed, and its path points and limit points in the task space are determined through linear trajectory planning. Then, the corresponding path points and limit points of each joint of the manipulator holding the laparoscope in the joint space are obtained by using the robotic kinematic model. Finally, the Chebyshev pseudo-spectral method and the sequential quadratic programming method are used to realize the trajectory planning of the robot joint space and the trajectory optimization based on the optimal time-smoothness. The laparoscopic following experiment shows that the maximum time of trajectory planning and optimization is 0.0362s, and the maximum speed and acceleration of the manipulator joint are no more than 1.68rad/s and 34.79rad/s2 respectively, which meets the requirements of laparoscopic real-time and smooth tracking of the distal movement of surgical instruments during laparoscopic minimally invasive surgery.
{"title":"Laparoscopic automatic following motion planning of minimally invasive surgery robot based on safety constraints","authors":"Shuizhong Zou, Yuan Huang, Ziang Wang","doi":"10.1109/ICMA54519.2022.9856048","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856048","url":null,"abstract":"In the master-slave operation stage of robot minimally invasive surgery, doctors need to frequently switch the control objects to adjust the position and posture of the laparoscope to obtain a better surgical field of vision, which will distract doctors' attention and lead to a decrease in the quality of laparoscopic surgery. Therefore, a laparoscopic automatic following motion planning is proposed to track the motion of surgical instruments quickly, smoothly and safely. Firstly, the position and posture adjustment conditions and adjusted visual field requirements of the distal reference point of the laparoscope are analyzed, and its path points and limit points in the task space are determined through linear trajectory planning. Then, the corresponding path points and limit points of each joint of the manipulator holding the laparoscope in the joint space are obtained by using the robotic kinematic model. Finally, the Chebyshev pseudo-spectral method and the sequential quadratic programming method are used to realize the trajectory planning of the robot joint space and the trajectory optimization based on the optimal time-smoothness. The laparoscopic following experiment shows that the maximum time of trajectory planning and optimization is 0.0362s, and the maximum speed and acceleration of the manipulator joint are no more than 1.68rad/s and 34.79rad/s2 respectively, which meets the requirements of laparoscopic real-time and smooth tracking of the distal movement of surgical instruments during laparoscopic minimally invasive surgery.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128787718","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 recent years, both convolution neural network and Transformer neural network have high popularity in the field of deep learning. These two kinds of neural networks have their own characteristics and are widely used in the field of speech recognition. Convolution neural network is good at dealing with local feature information, and the core module of Transformer is self-attention mechanism, so it has a good control of global information. In this paper, we combine these two kinds of networks, give full play to their respective advantages, use convolution neural network to extract the feature information from the spectrogram, and then give it to the Transformer network for global processing, so as to achieve a good recognition effect. End-to-end neural network often has some problems such as slow training speed and difficulty in training. in order to solve this problem, the spectrogram is used as the input of the network to reduce the amount of information processing of the network. on the other hand, the techniques such as batch normalization, layer normalization and residual network are applied in the model to speed up the training of the model and prevent the occurrence of over-fitting phenomenon.
{"title":"Speech recognition system of transformer improved by pre-parallel convolution Neural Network","authors":"Qi Yue, Zhang Han, Jing Chu, Xiaokai Han, Peiwen Li, Xuhui Deng","doi":"10.1109/ICMA54519.2022.9855999","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9855999","url":null,"abstract":"In recent years, both convolution neural network and Transformer neural network have high popularity in the field of deep learning. These two kinds of neural networks have their own characteristics and are widely used in the field of speech recognition. Convolution neural network is good at dealing with local feature information, and the core module of Transformer is self-attention mechanism, so it has a good control of global information. In this paper, we combine these two kinds of networks, give full play to their respective advantages, use convolution neural network to extract the feature information from the spectrogram, and then give it to the Transformer network for global processing, so as to achieve a good recognition effect. End-to-end neural network often has some problems such as slow training speed and difficulty in training. in order to solve this problem, the spectrogram is used as the input of the network to reduce the amount of information processing of the network. on the other hand, the techniques such as batch normalization, layer normalization and residual network are applied in the model to speed up the training of the model and prevent the occurrence of over-fitting phenomenon.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128275895","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}
Pub Date : 2022-08-07DOI: 10.1109/ICMA54519.2022.9856145
Yan Zhao, Yonglong Yu, Jun Kang, Yongcheng Zhang
To realize the monitoring and management of intelligent irrigation system, the NB-IOT-based intelligent monitoring and irrigation system for agriculture is designed. The system design adopts IOT architecture, the collection and control layer collects greenhouse environmental parameters through STM32 microcontroller, and transmits the data to the application layer through NB-IOT module. The system is tested by operation, and the intelligent monitoring and irrigation execution module can be adjusted and controlled according to the environmental and meteorological data uploaded by the NB-IOT module, so as to adapt to different weather and surrounding environment.
{"title":"Intelligent irrigation system based on NB-IOT","authors":"Yan Zhao, Yonglong Yu, Jun Kang, Yongcheng Zhang","doi":"10.1109/ICMA54519.2022.9856145","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856145","url":null,"abstract":"To realize the monitoring and management of intelligent irrigation system, the NB-IOT-based intelligent monitoring and irrigation system for agriculture is designed. The system design adopts IOT architecture, the collection and control layer collects greenhouse environmental parameters through STM32 microcontroller, and transmits the data to the application layer through NB-IOT module. The system is tested by operation, and the intelligent monitoring and irrigation execution module can be adjusted and controlled according to the environmental and meteorological data uploaded by the NB-IOT module, so as to adapt to different weather and surrounding environment.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128350162","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}
Pub Date : 2022-08-07DOI: 10.1109/ICMA54519.2022.9856105
Chenqi Li, Jian Guo, Shuxiang Guo, Qiang Fu
With the development of science and technology, many complex problems cannot be completed efficiently by a single robot and require multiple robots to work together. For complex task scenarios with multiple robots, the multi-robot task allocation problem is the key to coordinating robots to work efficiently. In this paper, for the application scenario of multi-robot collaborative inspection task allocation, the task allocation problem is first mathematically modelled using the multi-robot problem model, and simulations based on the resource balance search algorithm and the genetic population intelligence algorithm are applied respectively. Incorporating constraints in the population intelligence genetic algorithm, which transforms robot power constraints into distance constraints for research, allows for targeted simulation solutions for practical multi-robot collaborative detection. The results show that the genetic algorithm based on population intelligence can solve the problem well and minimise the total cost of multi-ball robot clustering, enhance the optimised search capability of the algorithm, improve the rationality of multi-robot task allocation and increase the efficiency of task completion.
{"title":"Study on Collaborative Task Assignment of Sphere Multi-Robot based on Group Intelligence Algorithm","authors":"Chenqi Li, Jian Guo, Shuxiang Guo, Qiang Fu","doi":"10.1109/ICMA54519.2022.9856105","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856105","url":null,"abstract":"With the development of science and technology, many complex problems cannot be completed efficiently by a single robot and require multiple robots to work together. For complex task scenarios with multiple robots, the multi-robot task allocation problem is the key to coordinating robots to work efficiently. In this paper, for the application scenario of multi-robot collaborative inspection task allocation, the task allocation problem is first mathematically modelled using the multi-robot problem model, and simulations based on the resource balance search algorithm and the genetic population intelligence algorithm are applied respectively. Incorporating constraints in the population intelligence genetic algorithm, which transforms robot power constraints into distance constraints for research, allows for targeted simulation solutions for practical multi-robot collaborative detection. The results show that the genetic algorithm based on population intelligence can solve the problem well and minimise the total cost of multi-ball robot clustering, enhance the optimised search capability of the algorithm, improve the rationality of multi-robot task allocation and increase the efficiency of task completion.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130648245","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}
Pub Date : 2022-08-07DOI: 10.1109/ICMA54519.2022.9856034
Xuesong Zhou, Donghui Xu, Youjie Ma, Fuhou Tu
With the rapid development of photovoltaic, wind power and other new energy sources, more and more renewable energy sources have been applied to the power market. However, there is still a serious phenomenon of “abandoning light and wind” at the present stage. In Addition, China’s fossil energy still occupies a dominant position. Therefore, how to improve the utilization rate of renewable energy and reduce carbon emissions is a problem worthy of study. Based on micro the renewable energy power suppliers and power grid in the net in the thermal power suppliers as the research object, based on cooperative game model is established and the consensus league chain mechanism, combined with the carbon emission reduction contribution and load characteristics, set up power and new energy power plant net income for the biggest target of cooperative game model, and through the analysis of the trading main body’s contribution to the alliance, Improve Shapley value profit distribution model. Simulation results show that: based on the cooperative game formed by blockchain, the total profit of power suppliers is higher than the sum of the profits of power suppliers under the traditional mode, which improves the utilization rate of renewable energy, reduces the rate of abandoned light and wind, and reduces the carbon emissions.
{"title":"Energy Block Chain Trading Mechanism Based on Game Strategy","authors":"Xuesong Zhou, Donghui Xu, Youjie Ma, Fuhou Tu","doi":"10.1109/ICMA54519.2022.9856034","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856034","url":null,"abstract":"With the rapid development of photovoltaic, wind power and other new energy sources, more and more renewable energy sources have been applied to the power market. However, there is still a serious phenomenon of “abandoning light and wind” at the present stage. In Addition, China’s fossil energy still occupies a dominant position. Therefore, how to improve the utilization rate of renewable energy and reduce carbon emissions is a problem worthy of study. Based on micro the renewable energy power suppliers and power grid in the net in the thermal power suppliers as the research object, based on cooperative game model is established and the consensus league chain mechanism, combined with the carbon emission reduction contribution and load characteristics, set up power and new energy power plant net income for the biggest target of cooperative game model, and through the analysis of the trading main body’s contribution to the alliance, Improve Shapley value profit distribution model. Simulation results show that: based on the cooperative game formed by blockchain, the total profit of power suppliers is higher than the sum of the profits of power suppliers under the traditional mode, which improves the utilization rate of renewable energy, reduces the rate of abandoned light and wind, and reduces the carbon emissions.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123861858","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}
Pub Date : 2022-08-07DOI: 10.1109/ICMA54519.2022.9856260
Ming Han, J. Sha, Yanheng Wang, C. Ma, Xiang Zhang
Detection or classification directly from 3D point clouds has received increasing attention in recent years. Transformer is more suitable for processing point cloud data than convolutional neural networks because of its inherent permutation invariance in processing sequences. However, common sampling strategies increase the training time of the model based on Transformer, such as Point Cloud Transformer (PCT). Aiming at the problem of slow inference speed of PCT, we propose a network structure named Fast Neighbor Embedding Point Cloud Transformer (FNE-PCT) in this paper. Instead of farthest point sample (FPS) and nearest neighbor search in PCT, FNE-PCT uses a fast neighbor embedding module to improve the inference speed and a residual self-attention encoding module to enhance the expression ability. Extensive experiments based on 3D object classification show that our FNE-PCT outperforms other excellent algorithms such as PointNet, PointNet++ and PointCNN. Our FNE-PCT achieves 92.6% accuracy on ModelNet40, which is on the same level as PCT. Meanwhile the speed is boosted up 29.2%, 43.6% and 52.9% than PCT respectively on ModelNet10, ModelNet40 and ShapeNetParts datasets.
{"title":"FNE-PCT: An Efficient Transformer Network for 3D Classification","authors":"Ming Han, J. Sha, Yanheng Wang, C. Ma, Xiang Zhang","doi":"10.1109/ICMA54519.2022.9856260","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856260","url":null,"abstract":"Detection or classification directly from 3D point clouds has received increasing attention in recent years. Transformer is more suitable for processing point cloud data than convolutional neural networks because of its inherent permutation invariance in processing sequences. However, common sampling strategies increase the training time of the model based on Transformer, such as Point Cloud Transformer (PCT). Aiming at the problem of slow inference speed of PCT, we propose a network structure named Fast Neighbor Embedding Point Cloud Transformer (FNE-PCT) in this paper. Instead of farthest point sample (FPS) and nearest neighbor search in PCT, FNE-PCT uses a fast neighbor embedding module to improve the inference speed and a residual self-attention encoding module to enhance the expression ability. Extensive experiments based on 3D object classification show that our FNE-PCT outperforms other excellent algorithms such as PointNet, PointNet++ and PointCNN. Our FNE-PCT achieves 92.6% accuracy on ModelNet40, which is on the same level as PCT. Meanwhile the speed is boosted up 29.2%, 43.6% and 52.9% than PCT respectively on ModelNet10, ModelNet40 and ShapeNetParts datasets.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121428266","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}
Pub Date : 2022-08-07DOI: 10.1109/ICMA54519.2022.9856066
Tao Huang, Xiaoyu Zou, Zhongbin Wang, Honglin Wu, Qingfeng Wang
Human detection within the operating area of electromechanical equipment is essential to ensure safe production and avoid accidents in the underground coal mine. Low light intensity and uneven light distribution in its environment surrenders the traditional color image based methods for human detection. In this paper, we focus on accurate detection of human in the operating area of electromechanical mining equipment using RGBD image. A novel network framework for miner detection based on YOLOv3 is proposed to fuse color image and depth image with enhanced attention mechanism. In the Pre-Backbone, feature extraction of both Depth and RGB branches are developed as the preliminary feature extractor with convolutional layer and residual block. Then the Convolutional Block Attention Module (CBAM) is improved to select and fuse RGB and Depth features by defining channel weights. Finally, the features are further inputted to Post-Backbone and used for multi-scale prediction in Head. The experiments demonstrate the superiority of the proposed method over some classical methods on miner detection with different light intensities and distributions.
{"title":"RGBD image based human detection for electromechanical equipment in underground coal mine","authors":"Tao Huang, Xiaoyu Zou, Zhongbin Wang, Honglin Wu, Qingfeng Wang","doi":"10.1109/ICMA54519.2022.9856066","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856066","url":null,"abstract":"Human detection within the operating area of electromechanical equipment is essential to ensure safe production and avoid accidents in the underground coal mine. Low light intensity and uneven light distribution in its environment surrenders the traditional color image based methods for human detection. In this paper, we focus on accurate detection of human in the operating area of electromechanical mining equipment using RGBD image. A novel network framework for miner detection based on YOLOv3 is proposed to fuse color image and depth image with enhanced attention mechanism. In the Pre-Backbone, feature extraction of both Depth and RGB branches are developed as the preliminary feature extractor with convolutional layer and residual block. Then the Convolutional Block Attention Module (CBAM) is improved to select and fuse RGB and Depth features by defining channel weights. Finally, the features are further inputted to Post-Backbone and used for multi-scale prediction in Head. The experiments demonstrate the superiority of the proposed method over some classical methods on miner detection with different light intensities and distributions.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114627081","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}
Pub Date : 2022-08-07DOI: 10.1109/ICMA54519.2022.9856254
Yuzhong Li, Niansong Zhang, Aiming Wang
Based on the demand for site optimization layout and scheduling proposed by the production plan of complex products based on the horizontal scientific research topic “Smart Site Scheduling and Monitoring Technology” of China Electric Power 14 Institute, the production efficiency of manufacturers and the reduction of scheduling are improved through reasonable site optimization layout scheduling technology. costs, so as to maximize the profit of the manufacturer.
{"title":"Layout and Scheduling Technology of Complex Product Production Site Based on Improved Genetic Algorithm and Placement Rules","authors":"Yuzhong Li, Niansong Zhang, Aiming Wang","doi":"10.1109/ICMA54519.2022.9856254","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856254","url":null,"abstract":"Based on the demand for site optimization layout and scheduling proposed by the production plan of complex products based on the horizontal scientific research topic “Smart Site Scheduling and Monitoring Technology” of China Electric Power 14 Institute, the production efficiency of manufacturers and the reduction of scheduling are improved through reasonable site optimization layout scheduling technology. costs, so as to maximize the profit of the manufacturer.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124158774","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}
Logistics delivery is widely used in many fields. This paper designs an intelligent logistics delivery experimental teaching system based on STM32F103 microcomputer and machine vision sensor. The convolutional neural network model based on improved LeNet-5 is applied to identify information of the destination number. The navigation information can be identified and traced at the same time. All information can be sent to the host computer through the serial port for processing. The experimental results show that the intelligent logistics delivery system can quickly complete the target number recognition and achieve high accuracy. The system has the advantages of low cost, moderate difficulty and stable performance. It has a certain value of experimental teaching and practical application.
{"title":"Design of Intelligent Logistics Delivery Experiment Teaching System based on Machine Vision","authors":"Xin Wang, Jintao Chen, Cong Li, Huiying Ma, Zhi Qi, Lihong Song","doi":"10.1109/ICMA54519.2022.9856082","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856082","url":null,"abstract":"Logistics delivery is widely used in many fields. This paper designs an intelligent logistics delivery experimental teaching system based on STM32F103 microcomputer and machine vision sensor. The convolutional neural network model based on improved LeNet-5 is applied to identify information of the destination number. The navigation information can be identified and traced at the same time. All information can be sent to the host computer through the serial port for processing. The experimental results show that the intelligent logistics delivery system can quickly complete the target number recognition and achieve high accuracy. The system has the advantages of low cost, moderate difficulty and stable performance. It has a certain value of experimental teaching and practical application.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127785515","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}
Pub Date : 2022-08-07DOI: 10.1109/ICMA54519.2022.9856086
Kai Wang, Jiwei Zhang, Hirotaka Hachiya, Haiyuan Wu
To interpret cardiac function through the use of echocardiography requires considerable expertise and years of diagnostic experience. To construct the support system for the evaluation of cardiac function from echocardiographic images, in this paper, we consider an automatic segmentation in a two-chamber view of echocardiographic images based on Attention U-Net. To improve accuracy, we made two ingenuity. 1) In the dataset, we merge the left ventricle as a medial constraint to its 6 parts of the left ventricular wall. 2) the weight of the corresponding loss function of each class is then set according to the area ratio of each class of echocardiography. Training and testing were performed using annotated data produced under the guidance of an echocardiographic expert.
{"title":"Study on Echocardiographic Image Segmentation Based on Attention U-Net","authors":"Kai Wang, Jiwei Zhang, Hirotaka Hachiya, Haiyuan Wu","doi":"10.1109/ICMA54519.2022.9856086","DOIUrl":"https://doi.org/10.1109/ICMA54519.2022.9856086","url":null,"abstract":"To interpret cardiac function through the use of echocardiography requires considerable expertise and years of diagnostic experience. To construct the support system for the evaluation of cardiac function from echocardiographic images, in this paper, we consider an automatic segmentation in a two-chamber view of echocardiographic images based on Attention U-Net. To improve accuracy, we made two ingenuity. 1) In the dataset, we merge the left ventricle as a medial constraint to its 6 parts of the left ventricular wall. 2) the weight of the corresponding loss function of each class is then set according to the area ratio of each class of echocardiography. Training and testing were performed using annotated data produced under the guidance of an echocardiographic expert.","PeriodicalId":120073,"journal":{"name":"2022 IEEE International Conference on Mechatronics and Automation (ICMA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126420261","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：