An Intelligent Tracking Technology Based on Kalman and Mean Shift Algorithm

Target tracking is one of the essential tasks in video-based surveillance system. A novel method which combined Kalman with Mean shift algorithm in moving object tracking is proposed in this paper. In the Mean Shift algorithm, a new adaptive bandwidth and the kernel weights have been adopted, meanwhile the extended Kalman filter has been used in fast moving object track. Experimental results show that the proposed method performs well in video-based target tracking.