REAL TIME TRAFFIC SURVEILLANCE VIDEO PROCESSING FOR VEHICLE TRAFFIC CONGESTION

Road Traffic has been one of the major concerns of a country as it affects many people both directly and indirectly. Metropolitan cities and traffic go hand in hand drawing inhabitants to congregate in large urban areas in intolerable levels of traffic congestion in urban areas and highways leading to cites. Effective urban governance requires a compromise between agglomeration and excessive congestion posing challenge for transportation, avoid unexpected congestion, follow the shortest path and less intersections to the destination based on historical data and the updates of real time traffic information. In order to obtain a statistical model of the flow of vehicles detailed monitoring of road traffic in intersections is extracted. In conventional methods, sensors at each of the intersection allow for counting, but are limited. The data is invaluable to understand the dynamic of a city's mobility working and improvement. New technique is developed to provide information that approaches to analyse video footage of intersections through computer vision algorithms to identify/track individual vehicles. The research aims to detect and track the vehicle from the video frame sequence, through vehicle motion along the frames using Background Subtraction Technique (BST), so that a statistical model may be extracted. The algorithm employs vehicle tracking with centroid based tracking by compensating for camera movement. Development of this approach proved useful for tracking cars, buses and trucks in highways encouraging for future expansion.