"DESIGN AND FPGA IMPLEMENTATION OF VIDEO IMAGE PROCESSOR FOR INTELLIGENT TRAFFIC CONTROL SYSTEM"

Abstract

In recent times, with continuously increasing vehicular densities, congestion at traffic intersections has become a major problem. Video image processing based intelligent traffic control systems have come to the fore to alleviate the problems associated with traffic congestion. However a major drawback is that real-time systems needed in processing video images require very high speed, which is not feasible using software environment in general purpose processors. Field Programmable Gate Array (FPGA) technology with their reconfigurable architecture offer to provide a viable solution for the implementation of algorithms suited to real time video image processing systems for traffic management applications. In this thesis, we propose a computationally efficient real-time video image processing system and its FPGA based implementation for traffic monitoring and control. The system is adaptive in nature and dynamically controls the traffic flow at intersections by taking into consideration the current traffic situation. It employs robust image processing algorithms to obtain information about optical flow from the input video image sequences. This information is used to control the duration of opening of a particular traffic channel