Real-time image tracking for automatic traffic monitoring and enforcement applications

This paper presents an image tracking system and its applications for traffic monitoring and accident detection at road intersections. Locations of motorcycles as well as automobiles are obtained in real time using the active contour model approach. Image measurement is further incorporated with Kalman filtering techniques to track individual vehicle motion. To initialize image tracking of vehicles at a junction, we propose a contour initialization method based on the concept of contour growing. Using a specially designed circuit board, a stand-alone image tracker has been designed and created for automatic traffic monitoring. We successfully achieved real-time image tracking of multi-lane vehicles. Interesting experimental results are presented to demonstrate the effectiveness of the proposed system. (C) 2003 Elsevier B.V. All rights reserved.