Application of machine vision to traffic monitoring and control

Research efforts have been made or are in progress on the application of machine vision to traffic monitoring and control. Of all the proposed systems, the wide-area detection system (WADS) appears to be the most reliable. However, limited field tests indicate that some malfunctions exist. Among these are false vehicle detection caused by sudden changes in road illumination and also by shadows or glare on the road produced by automobiles in adjacent lanes. After a brief review of WADS characteristics and problems, algorithm improvements to alleviate these problems are proposed. It is shown that the WADS vehicle detection algorithm becomes more reliable if the vehicle detection threshold includes a term depending on actual road intensity. Finally, an algorithm that would permit vehicle tracking for distances up to 120 m using available charge-coupled-device (CCD) cameras is proposed