Traffic Flow Surveillance System for Urban Intersections

This paper presents an automatic traffic surveillance system which is utilized to monitor the traffic intersections. Moving vehicles are extracted accurately from video sequences based on improved background modeling and object tracking method. In this paper, 16 types of general traffic behaviors at intersections are specifically defined and the traffic information collected from the tracking results is used to monitor them. The obtained information of the traffic flow at intersections can be utilized in intelligent traffic management to increase the efficiency and optimize spatial resources of urban road traffic infrastructures.