Improvement of Specific Multi-target Tracking Algorithm in Cross-domain Environment
[Purposes] Multi-target tracking in the cross-domain environment of surveillance video is a very important and challenging task in intelligent security. The difficulties of this task lie in the frequent occlusion between the objects in video frame, the unknown start and end time of the trajectory, t...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Editorial Office of Journal of Taiyuan University of Technology
2025-01-01
|
| Series: | Taiyuan Ligong Daxue xuebao |
| Subjects: | |
| Online Access: | https://tyutjournal.tyut.edu.cn/englishpaper/show-2376.html |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823856997022302208 |
|---|---|
| author | MU Xiaofang LI Hao LIU Jiaji LIU Zhenyu LI Yue |
| author_facet | MU Xiaofang LI Hao LIU Jiaji LIU Zhenyu LI Yue |
| author_sort | MU Xiaofang |
| collection | DOAJ |
| description | [Purposes] Multi-target tracking in the cross-domain environment of surveillance video is a very important and challenging task in intelligent security. The difficulties of this task lie in the frequent occlusion between the objects in video frame, the unknown start and end time of the trajectory, the too small sized targets, the interactions between the objects, the apparent similarity, and the camera angle changes. In view of the frequent occlusions and apparent similar problems, an improved multiple target tracking algorithm is put forward. [Methods] With the maximum use of low detection object, secondary matching is performed on the unmatched low objects. For target cross-domain, camera’s topological sort rules, adjacent cameras un tracking trajectory, as well as the detector YOLOv5 algorithm improvement and the layer-to-layer transfer of information streams, effectively address the multi-scale problems and the unsufficient information extraction problems for small objects, promptly match the tracking objects in adjacent camera, thus improving . the accuracy of multi-target tracking in cross-domain environment. [Findings] In the comparative ablation tests, the MOTA value of the improved algorithm reached 62.8%, and the IDswitch value was also significantly reduced. |
| format | Article |
| id | doaj-art-98a19c3f386245ea9b232047e670c9ba |
| institution | Kabale University |
| issn | 1007-9432 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Editorial Office of Journal of Taiyuan University of Technology |
| record_format | Article |
| series | Taiyuan Ligong Daxue xuebao |
| spelling | doaj-art-98a19c3f386245ea9b232047e670c9ba2025-02-12T03:34:23ZengEditorial Office of Journal of Taiyuan University of TechnologyTaiyuan Ligong Daxue xuebao1007-94322025-01-0156116517310.16355/j.tyut.1007-9432.202208431007-9432(2025)01-0165-09Improvement of Specific Multi-target Tracking Algorithm in Cross-domain EnvironmentMU Xiaofang0LI Hao1LIU Jiaji2LIU Zhenyu3LI Yue4College of Computer Science and Technology, Taiyuan Normal University, Jinzhong, Shanxi, ChinaCollege of Computer Science and Technology, Taiyuan Normal University, Jinzhong, Shanxi, ChinaCollege of Computer Science and Technology, Taiyuan Normal University, Jinzhong, Shanxi, ChinaCollege of Computer Science and Technology, Taiyuan Normal University, Jinzhong, Shanxi, ChinaSchool of Electronics and Information Engineering, Tiangong University, Tianjin, China[Purposes] Multi-target tracking in the cross-domain environment of surveillance video is a very important and challenging task in intelligent security. The difficulties of this task lie in the frequent occlusion between the objects in video frame, the unknown start and end time of the trajectory, the too small sized targets, the interactions between the objects, the apparent similarity, and the camera angle changes. In view of the frequent occlusions and apparent similar problems, an improved multiple target tracking algorithm is put forward. [Methods] With the maximum use of low detection object, secondary matching is performed on the unmatched low objects. For target cross-domain, camera’s topological sort rules, adjacent cameras un tracking trajectory, as well as the detector YOLOv5 algorithm improvement and the layer-to-layer transfer of information streams, effectively address the multi-scale problems and the unsufficient information extraction problems for small objects, promptly match the tracking objects in adjacent camera, thus improving . the accuracy of multi-target tracking in cross-domain environment. [Findings] In the comparative ablation tests, the MOTA value of the improved algorithm reached 62.8%, and the IDswitch value was also significantly reduced.https://tyutjournal.tyut.edu.cn/englishpaper/show-2376.htmlmulti-target trackingyolocomputer visiondeep learning |
| spellingShingle | MU Xiaofang LI Hao LIU Jiaji LIU Zhenyu LI Yue Improvement of Specific Multi-target Tracking Algorithm in Cross-domain Environment Taiyuan Ligong Daxue xuebao multi-target tracking yolo computer vision deep learning |
| title | Improvement of Specific Multi-target Tracking Algorithm in Cross-domain Environment |
| title_full | Improvement of Specific Multi-target Tracking Algorithm in Cross-domain Environment |
| title_fullStr | Improvement of Specific Multi-target Tracking Algorithm in Cross-domain Environment |
| title_full_unstemmed | Improvement of Specific Multi-target Tracking Algorithm in Cross-domain Environment |
| title_short | Improvement of Specific Multi-target Tracking Algorithm in Cross-domain Environment |
| title_sort | improvement of specific multi target tracking algorithm in cross domain environment |
| topic | multi-target tracking yolo computer vision deep learning |
| url | https://tyutjournal.tyut.edu.cn/englishpaper/show-2376.html |
| work_keys_str_mv | AT muxiaofang improvementofspecificmultitargettrackingalgorithmincrossdomainenvironment AT lihao improvementofspecificmultitargettrackingalgorithmincrossdomainenvironment AT liujiaji improvementofspecificmultitargettrackingalgorithmincrossdomainenvironment AT liuzhenyu improvementofspecificmultitargettrackingalgorithmincrossdomainenvironment AT liyue improvementofspecificmultitargettrackingalgorithmincrossdomainenvironment |