Challenges of Video-Based Vehicle Detection and Tracking in Intelligent Transportation Systems

Utilizing video-based instruments like surveillance cameras, shape the change in Intelligent Transportation Systems. These devices have attracted a great amount of research attention in traffic related applications due to their appropriate output quality, easy installation, low cost maintenance and high flexibility. The video output of cameras are considered as a valuable source of information in traffic control and management purposes. The developed video-based systems may use these video outputs to be analyzed for vehicle detection, counting, speed measurement and many other applications. Among them vehicle speed measurement using image processing techniques has become a common solution for speed violation detection and traffic control in recent decades. Video-based speed measurement is generally a sequential process consist of vehicle detection, tracking and relocation calculation in a certain time and region of the scene. This segmentation of tasks makes each step dependent to its previous phase’s accuracy. Therefor the false or inexact detection of vehicles may lead to incorrect tracking and then inaccurate relocation calculation which may produce error in measured speed value. This study is a review on the challenges of video-based speed measurement systems. Furthermore, taxonomy of the challenges and their causes are also discussed