Currently, many vision-based motion capture systems require passive markers attached to key loca- tions on the human body. However, such systems are intrusive with limited application. The algorithm that we use for hu...Currently, many vision-based motion capture systems require passive markers attached to key loca- tions on the human body. However, such systems are intrusive with limited application. The algorithm that we use for human motion capture in this paper is based on Markov random field (MRF) and dynamic graph cuts. It takes full account of the impact of 3D reconstruction error and integrates human motion capture and 3D reconstruction into MRF-MAP framework. For more accurate and robust performance, we extend our algorithm by incorporating color constraints into the pose estimation process. The advantages of incorporating color constraints are demonstrated by experimental results on several video sequences.展开更多
Aiming at the human–computer interaction control(HCIC)requirements of multi operators in collaborative virtual maintenance(CVM),real-time motion capture and simulation drive of multi operators with optical human moti...Aiming at the human–computer interaction control(HCIC)requirements of multi operators in collaborative virtual maintenance(CVM),real-time motion capture and simulation drive of multi operators with optical human motion capture system(HMCS)is proposed.The detailed realization process of real-time motion capture and data drive for virtual operators in CVM environment is presented to actualize the natural and online interactive operations.In order to ensure the cooperative and orderly interactions of virtual operators with the input operations of actual operators,collaborative HCIC model is established according to specific planning,allocating and decision-making of different maintenance tasks as well as the human–computer interaction features and collaborative maintenance operation features among multi maintenance trainees in CVM process.Finally,results of the experimental implementation validate the effectiveness and practicability of proposed methods,models,strategies and mechanisms.展开更多
As the effective capture region of optical motion capture system is limited by quantity,installation mode,resolution and focus of infrared cameras,the reflective markers on certain body parts(such as wrists,elbows,etc...As the effective capture region of optical motion capture system is limited by quantity,installation mode,resolution and focus of infrared cameras,the reflective markers on certain body parts(such as wrists,elbows,etc.)of multi-actual trainees may be obscured when they perform the collaborative interactive operation.To address this issue,motion data compensation method based on the additional feature information provided by the electromagnetic spatial position tracking equipment is proposed in this paper.The main working principle and detailed realization process of the proposed method are introduced step by step,and the practical implementation is presented to illustrate its validity and efficiency.The results show that the missing capture data and motion information of relevant obscured markers on arms can be retrieved with the proposed method,which can avoid the simulation motions of corresponding virtual operators being interrupted and deformed during the collaborative interactive operation process performed by multiactual trainees with optical human motion capture system in a limited capture range.展开更多
This study presents a kinematic calibration method for exoskeletal inertial motion capture (EI-MoCap) system with considering the random colored noise such as gyroscopic drift.In this method, the geometric parameters ...This study presents a kinematic calibration method for exoskeletal inertial motion capture (EI-MoCap) system with considering the random colored noise such as gyroscopic drift.In this method, the geometric parameters are calibrated by the traditional calibration method at first. Then, in order to calibrate the parameters affected by the random colored noise, the expectation maximization (EM) algorithm is introduced. Through the use of geometric parameters calibrated by the traditional calibration method, the iterations under the EM framework are decreased and the efficiency of the proposed method on embedded system is improved. The performance of the proposed kinematic calibration method is compared to the traditional calibration method. Furthermore, the feasibility of the proposed method is verified on the EI-MoCap system. The simulation and experiment demonstrate that the motion capture precision is significantly improved by 16.79%and 7.16%respectively in comparison to the traditional calibration method.展开更多
Recovering human pose from RGB images and videos has drawn increasing attention in recent years owing to minimum sensor requirements and applicability in diverse fields such as human-computer interaction,robotics,vide...Recovering human pose from RGB images and videos has drawn increasing attention in recent years owing to minimum sensor requirements and applicability in diverse fields such as human-computer interaction,robotics,video analytics,and augmented reality.Although a large amount of work has been devoted to this field,3D human pose estimation based on monocular images or videos remains a very challenging task due to a variety of difficulties such as depth ambiguities,occlusion,background clutters,and lack of training data.In this survey,we summarize recent advances in monocular 3D human pose estimation.We provide a general taxonomy to cover existing approaches and analyze their capabilities and limitations.We also present a summary of extensively used datasets and metrics,and provide a quantitative comparison of some representative methods.Finally,we conclude with a discussion on realistic challenges and open problems for future research directions.展开更多
基金Supported by the National Basic Research Program of China (Grant No.2006CB303105)
文摘Currently, many vision-based motion capture systems require passive markers attached to key loca- tions on the human body. However, such systems are intrusive with limited application. The algorithm that we use for human motion capture in this paper is based on Markov random field (MRF) and dynamic graph cuts. It takes full account of the impact of 3D reconstruction error and integrates human motion capture and 3D reconstruction into MRF-MAP framework. For more accurate and robust performance, we extend our algorithm by incorporating color constraints into the pose estimation process. The advantages of incorporating color constraints are demonstrated by experimental results on several video sequences.
文摘Aiming at the human–computer interaction control(HCIC)requirements of multi operators in collaborative virtual maintenance(CVM),real-time motion capture and simulation drive of multi operators with optical human motion capture system(HMCS)is proposed.The detailed realization process of real-time motion capture and data drive for virtual operators in CVM environment is presented to actualize the natural and online interactive operations.In order to ensure the cooperative and orderly interactions of virtual operators with the input operations of actual operators,collaborative HCIC model is established according to specific planning,allocating and decision-making of different maintenance tasks as well as the human–computer interaction features and collaborative maintenance operation features among multi maintenance trainees in CVM process.Finally,results of the experimental implementation validate the effectiveness and practicability of proposed methods,models,strategies and mechanisms.
基金the project supported by the National Natural Science Foundation of China(Grant No.61702524)the Natural Science Foundation of Shaanxi Province(Grant No.2016JQ6052).
文摘As the effective capture region of optical motion capture system is limited by quantity,installation mode,resolution and focus of infrared cameras,the reflective markers on certain body parts(such as wrists,elbows,etc.)of multi-actual trainees may be obscured when they perform the collaborative interactive operation.To address this issue,motion data compensation method based on the additional feature information provided by the electromagnetic spatial position tracking equipment is proposed in this paper.The main working principle and detailed realization process of the proposed method are introduced step by step,and the practical implementation is presented to illustrate its validity and efficiency.The results show that the missing capture data and motion information of relevant obscured markers on arms can be retrieved with the proposed method,which can avoid the simulation motions of corresponding virtual operators being interrupted and deformed during the collaborative interactive operation process performed by multiactual trainees with optical human motion capture system in a limited capture range.
基金supported by the National Natural Science Foundation o f China (61503392)。
文摘This study presents a kinematic calibration method for exoskeletal inertial motion capture (EI-MoCap) system with considering the random colored noise such as gyroscopic drift.In this method, the geometric parameters are calibrated by the traditional calibration method at first. Then, in order to calibrate the parameters affected by the random colored noise, the expectation maximization (EM) algorithm is introduced. Through the use of geometric parameters calibrated by the traditional calibration method, the iterations under the EM framework are decreased and the efficiency of the proposed method on embedded system is improved. The performance of the proposed kinematic calibration method is compared to the traditional calibration method. Furthermore, the feasibility of the proposed method is verified on the EI-MoCap system. The simulation and experiment demonstrate that the motion capture precision is significantly improved by 16.79%and 7.16%respectively in comparison to the traditional calibration method.
基金National Natural Science Foundation of China(61806176)the Fundamental Research Funds for the Central Universities(2019QNA5022).
文摘Recovering human pose from RGB images and videos has drawn increasing attention in recent years owing to minimum sensor requirements and applicability in diverse fields such as human-computer interaction,robotics,video analytics,and augmented reality.Although a large amount of work has been devoted to this field,3D human pose estimation based on monocular images or videos remains a very challenging task due to a variety of difficulties such as depth ambiguities,occlusion,background clutters,and lack of training data.In this survey,we summarize recent advances in monocular 3D human pose estimation.We provide a general taxonomy to cover existing approaches and analyze their capabilities and limitations.We also present a summary of extensively used datasets and metrics,and provide a quantitative comparison of some representative methods.Finally,we conclude with a discussion on realistic challenges and open problems for future research directions.