Interest points based object tracking with controlled cameras

A real time object tracking system based on multi-view cameras is proposed in this paper. In order to track a moving object, an active camera mounted on a pan/tilt platform controlled by static cameras is used in this system. Image mapping between the cameras in the system is described for the control. A background modelling based moving object detection and analysis model, especially a novel reliable and fast interest point detection technique for the real time object tracking system is proposed. Experiments to verify the method are presented. The system successfully tracks moving object from dynamic image with high noise background.