Real-time adaptive traffic light management system

Abstract

With an ever growing population and various modes of transport it has become vital to control and manage traffic. In many of the major countries across the world, roads have become one of the most popular means by which people and goods are transported from one place to another. However there is a limit to how much traffic roads can handle, which leads us to the problem of traffic flow and congestion within intersections and junctions. Computer vision is one of the most widely and actively researched areas in computer science. The reason for so much interest in this field comes from the concept of enabling computers to process and interpret images just like a human being would perceive the same images and to apply these computers in various fields, including video surveillance. When the potential of these computer systems were realized, they were applied to industrial needs with a degree of success. Traditionally, in order to monitor traffic, cameras were installed at various roads, from which personnel would have to spend hours on end looking at monitors. With the aid of computer vision, this process has been made easier by partially or sometimes fully replacing the personnel, which reduced cost and in some cases accuracy and efficiency. In this project a number of image process techniques were implemented in order to detect traffic densities in roads. This is the first component of the system. This component produces a scale of how dense the traffic is at a given time. The second component is a Fuzzy inference system which acts on the traffic densities in order to provide new timings for traffic lights to be adjusted accordingly. This component uses the values produced by the first component and acts on this value to produce new timings for the traffic lights in order to meliorate traffic flow hence minimize waiting time for cars. Results from the image processing component showed promising results and worked with a good degree of accuracy in most cases. The second component also showed that by acting on the traffic densities the new timings for traffic lights would lessen overall waiting times for cars (in certain instances) and improve traffic flow when compared to fixed traffic light timings in.