Adaptive collaborative tracking for multiple targets

作者：Quan Wei^*; Chen Jim X; Yu Nanyang

来源：Measurement Science and Technology, 2012, 23(12): 125407.

DOI：10.1088/0957-0233/23/12/125407

• multiple target tracking
• collaborative tracking
• motion correlation
• collaboration model
• occlusion
• target location estimation

This paper presents a new method for tracking multiple targets in video sequences. A common dilemma that multiple target tracking methods have to face in practice is reducing computational cost and addressing occlusion to achieve efficient and robust long-duration tracking. Existing methods mainly focus on deriving the measurement-to-track assignment through probabilistic analysis, in which the motions of targets are considered to be independent of each other. These methods tend to be computationally expensive due to the complexity of implementation for the visual tracking task. In this paper, we propose a collaboration model in which the acceleration difference between two targets is used to calculate the motion correlation value based on the two-dimensional Gaussian function. By the collaboration model, the location of occluded target is estimated using the motion information from other targets. In order to sense the occurrence of an occlusion accurately before estimating the location of target, we also propose a border-based occlusion decision method which is integrated into our tracking framework. The proposed approach is computationally efficient and robust. Experimental results exhibit the performance of the tracker based on our approach.

• 出版日期2012-12
• 单位西南交通大学