Improved optical flow estimation in traffic monitoring system
Huy-Hung Nguyen, Phuong-Tai Tran-Huu, Ha Manh Tran, Synh Viet‐Uyen Ha · 2013
Optical flow is the pattern that represents the motion of objects, edges, surfaces in real world by using displacement vectors or color flow. Application of optical flow is used in variant problems, especially in traffic monitoring system. The result of optical flow estimation can be used for traffic control, anomaly events detection, vehicle tracking and classification. In this paper, an approach that combines Brox's optical flow algorithm and background subtraction with a new method to improve the optical flow in the traffic monitoring system is presented. Our approach has demonstrated more accurate results than previous approaches.