A study on detection of vehicles' positions using HSV color transformed template images

Abstract

Position recognition of vehicles is important in ITS (intelligent transport system). This paper describes a vehicle position detection scheme for traffic flow images with a small depression angle. It uses template images divided into several color regions in the HSV (hue, saturation and value) color transform to reduce computational complexity. Finally, this proposed system has mostly performed in estimation of the vehicles' positions. However, the initial templates generated in the initial template generation module affects the detection performance.