Kalman filtering used in video-based traffic monitoring system

Video object tracking is an important method of traffic detection in Intelligent Transportation Systems. In video traffic tracking systems the matching method is often used to find the position of moving objects. In this article an improved algorithm of corner feature extraction is presented and corner points are tracked as the feature points of traffic objects. The tracking precision is mainly decided by matching algorithms. If the matching is not accurate, good tracking results cannot be achieved. In this article Kalman Filtering is used to track the moving traffic objects. In this system two kinds of data are used: One is from the general matching algorithm, which is the representation of the target's position; the other is detected by a spatial filtering velocimeter, containing the rough flow velocity of the targets. Though neither kind of data are highly accurate, Kalman Filtering is capable of integrating both position and velocity data to obtain better tracking results.