Vehicle detection based on spatial-temporal connection background subtraction

This paper describes an application of computer vision techniques to road surveillance. We could detect and track vehicles in real traffic scenes to generate meaningful traffic parameters as well as new metrics suitable for improved automated surveillance. This paper adopts spatial-temporal connection method to detect the vehicles. We use background subtraction based on Gaussian mixture modelling to extract the foreground, next update the foreground by spatial information. Experimental results and analysis of the algorithm are presented in this paper.