Computer Vision Based Approach for Overspeeding Problem in Smart Traffic System

With the growth in the urban population, count of vehicles on the road is also increasing drastically, traffic control in cities has become one of the most pressing challenges for the transportation system. A variety of different systems have been implemented across the country and around the world to resolve this issue. But most of them have proved inefficient to be implemented on a large scale that too in a developing country like India. Traffic management and related creative technologies are needed in the era of Machine Learning, Internet of Things (IoT), Image and Video processing, and Computer Vision in order to create more viable future cities. This paper presents a computer vision based approach for overspeed vehicle detection in Smart Traffic System (STS). Proposed overspeed vehicle detection system is based on centroid tracking and mark gap distance concept followed by OpenCV and Tesseract based method for license plate recognition. Primary purpose of the proposed system is to decrease cases of overspeeding and high death rates because of accidents. The accuracy of the proposed system is approximately 80% in detection of the overspeed vehicles.