Real-time pose estimation and motion tracking for motion performance using deep learning models
With the refinement and scientificization of sports training, the demand for sports performance analysis in the field of sports has gradually become prominent. In response to the problem of low accuracy and poor real-time performance in human pose estimation during sports, this article focused on vo...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
De Gruyter
2024-04-01
|
| Series: | Journal of Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1515/jisys-2023-0288 |
| _version_ | 1797197610948231168 |
|---|---|
| author | Liu Long Dai Yuxin Liu Zhihao |
| author_facet | Liu Long Dai Yuxin Liu Zhihao |
| author_sort | Liu Long |
| collection | DOAJ |
| description | With the refinement and scientificization of sports training, the demand for sports performance analysis in the field of sports has gradually become prominent. In response to the problem of low accuracy and poor real-time performance in human pose estimation during sports, this article focused on volleyball sports and used a combination model of OpenPose and DeepSORT to perform real-time pose estimation and tracking on volleyball videos. First, the OpenPose algorithm was adopted to estimate the posture of the human body region, accurately estimating the coordinates of key points, and assisting the model in understanding the posture. Then, the DeepSORT model target tracking algorithm was utilized to track the detected human pose information in real-time, ensuring consistency of identification and continuity of position between different frames. Finally, using unmanned aerial vehicles as carriers, the YOLOv4 object detection model was used to perform real-time human pose detection on standardized images. The experimental results on the Volleyball Activity Dataset showed that the OpenPose model had a pose estimation accuracy of 98.23%, which was 6.17% higher than the PoseNet model. The overall processing speed reached 16.7 frames/s. It has good pose recognition accuracy and real-time performance and can adapt to various volleyball match scenes. |
| first_indexed | 2024-04-24T06:46:43Z |
| format | Article |
| id | doaj.art-87e16281cd2444b885e6dfab853842e5 |
| institution | Directory Open Access Journal |
| issn | 2191-026X |
| language | English |
| last_indexed | 2024-04-24T06:46:43Z |
| publishDate | 2024-04-01 |
| publisher | De Gruyter |
| record_format | Article |
| series | Journal of Intelligent Systems |
| spelling | doaj.art-87e16281cd2444b885e6dfab853842e52024-04-22T19:40:04ZengDe GruyterJournal of Intelligent Systems2191-026X2024-04-0133113710.1515/jisys-2023-0288Real-time pose estimation and motion tracking for motion performance using deep learning modelsLiu Long0Dai Yuxin1Liu Zhihao2School of Health Care, Chongqing Preschool Education College, Chongqing404047, ChinaSchool of Physical Education, Chongqing Preschool Education College, Chongqing404047, ChinaSchool of Physical Education, Chongqing Preschool Education College, Chongqing404047, ChinaWith the refinement and scientificization of sports training, the demand for sports performance analysis in the field of sports has gradually become prominent. In response to the problem of low accuracy and poor real-time performance in human pose estimation during sports, this article focused on volleyball sports and used a combination model of OpenPose and DeepSORT to perform real-time pose estimation and tracking on volleyball videos. First, the OpenPose algorithm was adopted to estimate the posture of the human body region, accurately estimating the coordinates of key points, and assisting the model in understanding the posture. Then, the DeepSORT model target tracking algorithm was utilized to track the detected human pose information in real-time, ensuring consistency of identification and continuity of position between different frames. Finally, using unmanned aerial vehicles as carriers, the YOLOv4 object detection model was used to perform real-time human pose detection on standardized images. The experimental results on the Volleyball Activity Dataset showed that the OpenPose model had a pose estimation accuracy of 98.23%, which was 6.17% higher than the PoseNet model. The overall processing speed reached 16.7 frames/s. It has good pose recognition accuracy and real-time performance and can adapt to various volleyball match scenes.https://doi.org/10.1515/jisys-2023-0288motion performancepose estimationmotion trackingdeep learning modelsreal-time performance |
| spellingShingle | Liu Long Dai Yuxin Liu Zhihao Real-time pose estimation and motion tracking for motion performance using deep learning models Journal of Intelligent Systems motion performance pose estimation motion tracking deep learning models real-time performance |
| title | Real-time pose estimation and motion tracking for motion performance using deep learning models |
| title_full | Real-time pose estimation and motion tracking for motion performance using deep learning models |
| title_fullStr | Real-time pose estimation and motion tracking for motion performance using deep learning models |
| title_full_unstemmed | Real-time pose estimation and motion tracking for motion performance using deep learning models |
| title_short | Real-time pose estimation and motion tracking for motion performance using deep learning models |
| title_sort | real time pose estimation and motion tracking for motion performance using deep learning models |
| topic | motion performance pose estimation motion tracking deep learning models real-time performance |
| url | https://doi.org/10.1515/jisys-2023-0288 |
| work_keys_str_mv | AT liulong realtimeposeestimationandmotiontrackingformotionperformanceusingdeeplearningmodels AT daiyuxin realtimeposeestimationandmotiontrackingformotionperformanceusingdeeplearningmodels AT liuzhihao realtimeposeestimationandmotiontrackingformotionperformanceusingdeeplearningmodels |