Laser & Optoelectronics Progress, Volume. 59, Issue 2, 0215006(2022)
Multi-Target Tracking Algorithm Based on Improved Markov Random Field Model of Optical Flow Motion Constraint
Figures & Tables(7)
Fig. 4. Tracking results of sequence KITTI-17. (a) 1st frame; (b) 8th frame; (c) 14th frame;
Fig. 5. Tracking results of sequence PETS09-S2L1. (a) 430th frame; (b) 462nd frame; (c) 474th frame;
Fig. 6. Tracking results of sequence KITTI-TEST17. (a) 9th frame; (b) 22nd frame; (c) 37th frame; (d) 56th frame;
Table 1. Performance comparison of different algorithms
View tableTable 1. Performance comparison of different algorithms
Sequence Algorithm MOTA /% MOTP /% IDF1 /% FN FP IDSW KITTI-17 Algorithm in reference ［ 20 ］65.2 72.7 78.2 130 32 13 Algorithm in reference ［ 21 ］77.6 81.3 71.3 89 44 20 Algorithm in reference ［ 22 ］79.8 82.1 147 53 32 Proposed algorithm 83.9 87.6 80.7 96 38 6 PETS09-S2L1 Algorithm in reference ［ 20 ］57.6 65.5 76.3 609 265 24 Algorithm in reference ［ 21 ］85.3 82.8 79.6 279 185 10 Algorithm in reference ［ 22 ］87.4 76.6 400 123 62 Proposed algorithm 89.6 84.4 83.0 232 156 9
Tools
Get Citation
Copy Citation Text
Yuhui Niu, Zhenghao Xi, Yajing Xue, Jianchao Chen. Multi-Target Tracking Algorithm Based on Improved Markov Random Field Model of Optical Flow Motion Constraint[J]. Laser & Optoelectronics Progress, 2022, 59(2): 0215006
Paper Information
Category: Machine Vision
Received: Mar. 4, 2021
Accepted: Apr. 14, 2021
Published Online: Dec. 29, 2021
The Author Email: Yuhui Niu (n18321023297@163.com), Yajing Xue (xyj2012yjs@163.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20