Real-time multi-person tracking in fixed surveillance camera environment

Abstract

In this paper, we propose a real-time multi-person tracking system operating in fixed surveillance camera environment. We adopt particle filtering as our object tracking framework. Background subtraction is used to generate the ROI. And pedestrian detection is used to initialize each tracker. Object size estimation and tracking failure detection is proposed to improve tracking accuracy and robustness. Experimental results demonstrate that the proposed algorithm tracks multiple persons efficiently in real-time.