Multi-Target Tracking with Occlusion Handling: A Fusion of Mean Shift and Unscented Kalman Filtering
This paper proposes a multi-target tracking method that combines the mean shift algorithm and Unscented Kalman Filter (UKF). When a spatial target is temporarily occluded, the trajectory is predicted by Unscented Kalman Filter, and the trajectories of the detected and predicted targets are correlated using the Hungarian algorithm to achieve efficient tracking. The experimental results show that the improved algorithm is not only able to track the temporarily occluded target, but also has a MOTPA value of 91.754% and a MOTP of 93.237%.
Authors
- Xu Yang (ORCID: https://orcid.org/0000-0003-0553-4581)
- Bingchuan Li
- Chenggang Dai (ORCID: https://orcid.org/0000-0002-8386-794X)
- Zhengxu Zhao (ORCID: https://orcid.org/0000-0002-6066-623X)
- Jingqi Yang (ORCID: https://orcid.org/0000-0002-5762-5031)
- Qinghao Pang (ORCID: https://orcid.org/0009-0008-2762-4557)
Institutions
- Chinese Academy of Sciences (CN)
- National Astronomical Observatories (CN)
- Qingdao University of Technology (CN)
Publication Details
- Published
- 2024-10-25
- DOI
- https://doi.org/10.1109/isstc63573.2024.10824177
- Primary Topic
- Target Tracking and Data Fusion in Sensor Networks
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Qingdao University