Sensor signal processing for IVHS applications

Accurate and wide-area estimates of vehicle velocity and traffic spatial and temporal densities will be essential components of future algorithms for freeway and arterial street control, for incident prediction and detection, and for optimization in route selection. Algorithms like these figure prominently in the current research and development of intelligent vehicle-highway systems (IVHS). This paper presents an approach to a class of vehicle monitoring problems which is based upon a video backbone sensor and multiple target tracking (MTT). The method allows the integration of measurements made from other sensors like inductive loops, microwave radars, and laser range profilers