In present-day industrial settings,where robot arms performtasks in an unstructured environment,theremay exist numerousobjects of various shapes scattered in randompositions,making it challenging for a robot armtoprec...In present-day industrial settings,where robot arms performtasks in an unstructured environment,theremay exist numerousobjects of various shapes scattered in randompositions,making it challenging for a robot armtoprecisely attain the ideal pose to grasp the object.To solve this problem,a multistage robotic arm flexible grasp detection method based on deep learning is proposed.This method first improves the Faster RCNN target detection model,which significantly improves the detection ability of the model for multiscale grasped objects in unstructured scenes.Then,a Squeeze-and-Excitation module is introduced to design a multitarget grasping pose generation network based on a deep convolutional neural network to generate a variety of graspable poses for grasped objects.Finally,a multiobjective IOU mixed area attitude evaluation algorithm is constructed to screen out the optimal grasping area of the grasped object and obtain the optimal grasping posture of the robotic arm.The experimental results show that the accuracy of the target detection network improved by the method proposed in this paper reaches 96.6%,the grasping frame accuracy of the grasping pose generation network reaches 94%and the flexible grasping task of the robotic arm in an unstructured scene in a real environment can be efficiently and accurately implemented.展开更多
Detecting feature points on the human body in video frames is a key step for tracking human movements. There have been methods developed that leverage models of human pose and classification of pixels of the body imag...Detecting feature points on the human body in video frames is a key step for tracking human movements. There have been methods developed that leverage models of human pose and classification of pixels of the body image. Yet, occlusion and robustness are still open challenges. In this paper, we present an automatic, model-free feature point detection and action tracking method using a time-of-flight camera. Our method automatically detects feature points for movement abstraction. To overcome errors caused by miss-detection and occlusion, a refinement method is devised that uses the trajectory of the feature points to correct the erroneous detections. Experiments were conducted using videos acquired with a Microsoft Kinect camera and a publicly available video set and comparisons were conducted with the state-of-the-art methods. The results demonstrated that our proposed method delivered improved and reliable performance with an average accuracy in the range of 90 %.The trajectorybased refinement also demonstrated satisfactory effectiveness that recovers the detection with a success rate of 93.7 %. Our method processed a frame in an average time of 71.1 ms.展开更多
With the rapid advancements in artificial intelligence and computer vision technology,thefield of visual-based human pose detection has emerged as a highly sought-after research area in recent years.The identification of...With the rapid advancements in artificial intelligence and computer vision technology,thefield of visual-based human pose detection has emerged as a highly sought-after research area in recent years.The identification of human poses has practical applications in diverse domains,ranging from motion-sensing games for human-computer interaction to activity prediction and medical rehabil-itation.The present study is focused on the utilization of human pose detection forfitness movement counting.The ultimate aim of the system design is to accurately detect the skeletal key points of each body part in the image and subsequently con-nect them to form a human pose skeleton,which serves as a vital representation of the characteristics of human motion,particularly in the context of video data,where multiple human poses can be linked to form a certain movement trajectory.By judging the trajectory and angle changes,the system can determine whether people’sfitness movements are correct and help them improve theirfitness effec-tiveness.Hence,an increasing number of researchers are investing time and effort in thisfield.One common approach for human pose detection is OpenPose,but this model has a large and complex structure and low detection accuracy.Therefore,thisfitness movement detection and counting system uses a lightweight MediaPipe model and improves it to enhance the algorithm’s accuracy and recognition speed.The specific work in this paper includes three main points:(1)a suitable network structure to detect human skeletal points;(2)the appropriate skeletal structure forfitness movements through experiments to obtain accurate results;and(3)a Qt interface for human-computer interaction.展开更多
Supported by the Science Fund of the Creative Research Group,the research team led by Prof.Chen Hualan(陈化兰)in Harbin Veterinary Research Institute,Chinese Academy of Agricultural Sciences found that the low pathoge...Supported by the Science Fund of the Creative Research Group,the research team led by Prof.Chen Hualan(陈化兰)in Harbin Veterinary Research Institute,Chinese Academy of Agricultural Sciences found that the low pathogenic H7N9viruses emerging in 2013have mutated to highly pathogenic viruses in chickens and are more dangerous to humans,which was published in Cell Research(2017,doi:10.1038/cr.2017.129).展开更多
基金supported in part by the National Natural Science Foundation of China(No.52165063)Guizhou Provincial Science and Technology Projects(Qiankehepingtai-GCC[2022]006-1,Qiankehezhicheng[2021]172,[2021]397,[2021]445,[2022]008,[2022]165)+1 种基金Natural Science Research Project of Guizhou Provincial Department of Education(Qianjiaoji[2022]No.436)Guizhou Province Graduate Research Fund(YJSCXJH[2021]068).
文摘In present-day industrial settings,where robot arms performtasks in an unstructured environment,theremay exist numerousobjects of various shapes scattered in randompositions,making it challenging for a robot armtoprecisely attain the ideal pose to grasp the object.To solve this problem,a multistage robotic arm flexible grasp detection method based on deep learning is proposed.This method first improves the Faster RCNN target detection model,which significantly improves the detection ability of the model for multiscale grasped objects in unstructured scenes.Then,a Squeeze-and-Excitation module is introduced to design a multitarget grasping pose generation network based on a deep convolutional neural network to generate a variety of graspable poses for grasped objects.Finally,a multiobjective IOU mixed area attitude evaluation algorithm is constructed to screen out the optimal grasping area of the grasped object and obtain the optimal grasping posture of the robotic arm.The experimental results show that the accuracy of the target detection network improved by the method proposed in this paper reaches 96.6%,the grasping frame accuracy of the grasping pose generation network reaches 94%and the flexible grasping task of the robotic arm in an unstructured scene in a real environment can be efficiently and accurately implemented.
文摘Detecting feature points on the human body in video frames is a key step for tracking human movements. There have been methods developed that leverage models of human pose and classification of pixels of the body image. Yet, occlusion and robustness are still open challenges. In this paper, we present an automatic, model-free feature point detection and action tracking method using a time-of-flight camera. Our method automatically detects feature points for movement abstraction. To overcome errors caused by miss-detection and occlusion, a refinement method is devised that uses the trajectory of the feature points to correct the erroneous detections. Experiments were conducted using videos acquired with a Microsoft Kinect camera and a publicly available video set and comparisons were conducted with the state-of-the-art methods. The results demonstrated that our proposed method delivered improved and reliable performance with an average accuracy in the range of 90 %.The trajectorybased refinement also demonstrated satisfactory effectiveness that recovers the detection with a success rate of 93.7 %. Our method processed a frame in an average time of 71.1 ms.
文摘With the rapid advancements in artificial intelligence and computer vision technology,thefield of visual-based human pose detection has emerged as a highly sought-after research area in recent years.The identification of human poses has practical applications in diverse domains,ranging from motion-sensing games for human-computer interaction to activity prediction and medical rehabil-itation.The present study is focused on the utilization of human pose detection forfitness movement counting.The ultimate aim of the system design is to accurately detect the skeletal key points of each body part in the image and subsequently con-nect them to form a human pose skeleton,which serves as a vital representation of the characteristics of human motion,particularly in the context of video data,where multiple human poses can be linked to form a certain movement trajectory.By judging the trajectory and angle changes,the system can determine whether people’sfitness movements are correct and help them improve theirfitness effec-tiveness.Hence,an increasing number of researchers are investing time and effort in thisfield.One common approach for human pose detection is OpenPose,but this model has a large and complex structure and low detection accuracy.Therefore,thisfitness movement detection and counting system uses a lightweight MediaPipe model and improves it to enhance the algorithm’s accuracy and recognition speed.The specific work in this paper includes three main points:(1)a suitable network structure to detect human skeletal points;(2)the appropriate skeletal structure forfitness movements through experiments to obtain accurate results;and(3)a Qt interface for human-computer interaction.
文摘Supported by the Science Fund of the Creative Research Group,the research team led by Prof.Chen Hualan(陈化兰)in Harbin Veterinary Research Institute,Chinese Academy of Agricultural Sciences found that the low pathogenic H7N9viruses emerging in 2013have mutated to highly pathogenic viruses in chickens and are more dangerous to humans,which was published in Cell Research(2017,doi:10.1038/cr.2017.129).