Pixel detection and elimination algorithm to control traffic congestion aided by Fuzzy logic

In the modern era, the escalation of vehicles on the roads has caused an increasing need for a reliable and intelligent control of the traffic light system. This paper is aimed at solving this crisis by effectively computing the density of traffic based on the images picked up by cameras placed on the traffic posts. The method involves a simple algorithm which performs pixel elimination and detection followed by processing using a fuzzy controller. It can be further extended towards hardware implementation using dedicated processors. Our approach involves taking images at regular intervals and continuously processing them with a reference image which is captured when there is no traffic (empty road).The reference images are stored and used for calibration purpose. This algorithm enables the system to infer the traffic density which is then evaluated by a fuzzy controller to determine the timing of the traffic signals. The fuzzy controller consists of an output function which dynamically controls the output based on the comparison of current image's pixel count corresponding to the vehicle density.