Moving object tracking system based on camshift and Kalman filter

To accomplish real-time tracking of moving objects requirements, and overcome the defect of occlusion in the process of tracking moving object, this paper presents a set of real-time tracking system. The tracking system uses a combination of camshift and kalman filter algorithm. When the moving object is a large area blocked, the velocity of moving object is applied linear prediction to kalman filter tracking. When the camera moves, we can quickly find the target in this tracking system. The experimental results show that : this system designed for tracking moving targets has good robustness.