基于Camshift和Kalman滤波结合的改进多目标跟踪算法

Improved Tracking Algorithm for Multiple Targets Based on Camshift Algorithm Combined with Kalman Filter

Camshift算法具有很好的实时性,但是存在遇到大面积与目标颜色相近的背景干扰、目标被严重遮挡时跟踪失效,只能对单目标进行跟踪等问题。针对这些问题,提出了Camshift结合Kalman滤波的改进算法,采用改进算法的跟踪器对每个检测出来的目标进行分别跟踪,从而实现多目标跟踪。实验表明本文提出的改进算法能有效地克服以上问题,并且能达到实时性要求。

Camshift algorithm has the advantage of better real-time,but the algorithm will track failure when large-area background color is similar to object object is occlusioned seriously,and is limited tracking single target,etc.Aim-ing at these problems,this paper presents the improved method of multiple targets tracking algorithm based on Camshift algorithm combined with Kalman fi lter.The tracker of the improved method was used to track each detected target,it can achieve tracking of multi-targets.The results showed that the proposed algorithm can overcome the above problems effec-tively and achieve real-time requirements.