Share Email Print

Optical Engineering

Online two-stage association method for robust multiple people tracking
Author(s): Jingqin Lv; Jiangxiong Fang; Jie Yang
Format Member Price Non-Member Price
PDF $20.00 $25.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Robust multiple people tracking is very important for many applications. It is a challenging problem due to occlusion and interaction in crowded scenarios. This paper proposes an online two-stage association method for robust multiple people tracking. In the first stage, short tracklets generated by linking people detection responses grow longer by particle filter based tracking, with detection confidence embedded into the observation model. And, an examining scheme runs at each frame for the reliability of tracking. In the second stage, multiple people tracking is achieved by linking tracklets to generate trajectories. An online tracklet association method is proposed to solve the linking problem, which allows applications in time-critical scenarios. This method is evaluated on the popular CAVIAR dataset. The experimental results show that our two-stage method is robust.

Paper Details

Date Published: 1 July 2011
PDF: 6 pages
Opt. Eng. 50(7) 077207 doi: 10.1117/1.3600423
Published in: Optical Engineering Volume 50, Issue 7
Show Author Affiliations
Jingqin Lv, Shanghai Jiao Tong Univ. (China)
Jiangxiong Fang, Shanghai Jiao Tong Univ. (China)
Jie Yang, Shanghai Jiao Tong Univ. (China)

© SPIE. Terms of Use
Back to Top