Real-time image tracking for traffic monitoring

This paper presents an image tracking system and its application to traffic parameter estimation at a road intersection. An active contour model is adopted to obtain vehicle locations of motorcycles as well as automobiles. Image location measurement is further incorporated with Kalman filter techniques to track the individual vehicle motion. To initialize image tracking at a junction, we propose a design based on the concept of contour growing. Traffic parameters such as vehicle speed and numbers of vehicles can be calculated with satisfactory accuracy. Actual CCD camera images of a busy junction were employed to verify the proposed method.