Intelligent Traffic Control System Using Image Processing

Abstract

Traffic congestion is a problem in all developing and developed countries around the world.
Thus, efforts such as controlling the traffic by integrating sensors to the traffic light system have been
carried out, but the problem persists. Therefore, to reduce the queue at traffic lights, an image processingbased solution is proposed in this paper. The proposed system uses the Region-Convolutional Neural
Network to measure vehicle density and the Optical Character Recognition (OCR) to detect ambulances.
The proposed system was able to detect traffic with at least 80% accuracy and an ambulance with almost
100% accuracy. This information could be used to better manage traffic.