A window-based image processing approach for measuring traffic movements at junctions

Road traffic data collection and analysis is extensively required by the traffic engineers for traffic control and management. Image processing techniques with wide area sensing and analysis capabilities have several advantages over the conventional point based sensors, like magnetic loops etc. So far, many different algorithms have been proposed for measuring simple traffic parameters, but less attention has been paid to measure complex traffic parameters such as traffic movements at junctions (TMJ). In this paper, we described a low-cost image processing approach for measuring the TMJ parameters in real-time. An effective edge detection technique is used to detect the vehicles and to measure the TMJ parameters in real-time.