Detection and tracking of multiple moving targets in complex conditions

DOI

10.2991/icmmcce-15.2015.577How to use a DOI?

Keywords

scene constraint, motion estimation, particle filter, MCMC, multiple target tracking.

Abstract

Detection and tracking of multiple moving targets in complex conditions has always been a hot and difficult problem, which mainly includes that the tracked target is occluded, the characteristics of target and background are similar, illumination and background are changing over time, etc. In order to solve this problem, a method of multiple moving targets detection and tracking is proposed in this paper, which can be applied to complex conditions. In the aspect of target detection, on the basis of mixed multi-scales and deformable model, we have proposed scene constraint and the target own property to reduce the potential areas of targets and search scope and to filter interference object, which has improved the efficiency of detection algorithm greatly. In the aspect of target tracking, we use global motion estimation to modify motion model of tracking algorithm, at the same time, propose an improved particle filter algorithm which has introduced reversibility-jump mechanism and joined MCMC improvement strategy. By building the observation model with the color histogram of target RGB, it can track the moving targets in different conditions accurately. The experimental results show that the method proposed in this paper can accurately track the multiple moving targets in complex conditions, and the effect is significant.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Tao Xu
AU  - Chaobing Huang
AU  - Mengqiang Zhu
PY  - 2015/12
DA  - 2015/12
TI  - Detection and tracking of multiple moving targets in complex conditions
BT  - Proceedings of the 4th International Conference on Mechatronics, Materials, Chemistry and Computer Engineering 2015
PB  - Atlantis Press
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmmcce-15.2015.577
DO  - 10.2991/icmmcce-15.2015.577
ID  - Xu2015/12
ER  -