Target tracking based on the improved Camshift method
Chunbo Xiu, Fushan Ba · 2016
In target tracking, the complex background often has a negative effect on the tracking quality. The feature extraction of the moving target can be extracted in order to suppress the interference caused by background tracking. An improved Camshift tracking method is proposed in this paper, a gauss weight function is chosen to select the target area that is called the region of interest and intercept the region of interest from the background, the back projection map of the target area is tracked independently, thus eliminating the interference of the background to the object. The proposed method is compared with the traditional Camshift algorithm and the Camshift algorithm based on multi feature fusion, the experimental results show that the proposed Camshift algorithm based on multi feature fusion has higher accuracy and stability compared with the traditional Camshift tracking algorithm. It overcomes the problem of interference caused by the complex background in the target tracking. It has high practicability and meets the requirement of real-time.