Review on vehicle detection based on video for traffic surveillance

Abstract

Developing real-time traffic parameters surveillance systems based on video aiming to extract reliable traffic state information has attracted a lot of attention during the past decades. These traffic state parameters include traffic flow density, the length of queue, average traffic speed and total vehicle in fixed time intervals. In these systems, robust and reliable vehicle detection and tracking technique is a critical step. This paper presents a review of recent vehicle detection and tracking techniques based on stationary video. Our focus is on systems where the rectilinear stationary camera is configured on roadside rather than mounted on the vehicle. Generally, three steps are utilized to acquire traffic condition information, including selection of region of interest (ROI), vehicle detection and vehicle tracking. First, we discuss various methods related to the selection of region of interest (ROI) by a brief review. Then a variety of algorithms and techniques developed to detect vehicle are discussed. Integrating vehicle detection methods are also illustrated to extract robust and accurate traffic parameters. Finally, a brief overview on shadow eliminated techniques is also presented.