Image Processing in Intelligent Traffic Management

Abstract

Traffic monitoring and traffic control have always been challenging tasks. Intelligent Transportation Systems (ITS) based on wide range of technologies have certain practical challenges in their application and implementation. Video surveillance has proven advantageous over traditional systems based on inductive loops sensors and detectors for traffic monitoring. Accurate traffic density estimation which is basic to tackling traffic congestions requires detection of vehicles, assessing their speed, and tracking vehicles passing through surveillance zones. Image processing techniques require processing of large number of image frames for real-time applications in traffic management. More efficient and less costly image processing techniques for accurate vehicle detection and density determination are required for developing more effective traffic management systems. There is a need for developing algorithms with robust performance under heavy traffic loads and varied environmental conditions. Developments in artificial intelligence offer new vistas in image processing for regulation and management of traffic by signal control mechanisms and creation of neural networks for unhindered traffic flow.