Traffic Violation using Distance Estimation and CNN

There has been an extensive need for an efficient model to approximate distances between objects on real time video surveillance. This paper presents a novel distance estimation to calculate distances from the stereo camera. The distance between contours of the vehicles is found using centroid formula and Euclid’s estimation. The end product is a Machine Learning algorithm that provides an efficient solution to monitor traffic violations as well as swiftly catch the perpetrators on real time video.