Traffic Analysis and Signal Timer Management Using OpenCV

Abstract

Traffic has been a major issue at intersections throughout the world, here a technique is proposed to reduce manpower required to handle the traffic and remove static timers from intersections. Proposed system is consisting of simple computer device with CCTV. It works by analyse traffic condition from video input and then count vehicles to manipulate signal timer which avoid traffic collision and maintain traffic flow. Therefor the first step in this process is taking video input from mounted CCTV camera and then detection of cars. The system uses Haar like features, which is mainly made for face detection. Haar feature-based cascade classifier is effective object detection technique. It has data set containing positive and negative image data which help agent to identify target object. Result shows this system is more effective in detection of cars compare to existing systems.