Automated Traffic Detection System Based on Image Processing

Abstract

This paper proposes a low-cost automated traffic detection system based on image processing. Dhaka is one of the crowded cities in the world with highly challenging traffic system. There is substantial lack of awareness among the drivers of transport system. As a result, citizens do not follow the rules and regulation while driving in Dhaka city. The tendency of violating the traffic regulation is noticeable throughout the country. As a result, the whole traffic system collapses very often and sometimes it ends-up with severe accidents. In recent days, the government has taken different initiatives including enlargement of pedestrian walkways, building new flyovers and foot-over bridges, expansion of existing roads. But, violation still the outcome of all these initiatives could not improve the situation significantly. The proposed system will automatically detect the traffic through live streaming video so that the detected images can be used to detect traffic violation. Later on, the law enforcement agency will be able to take necessary legal steps based on the stored information on the database.