A rough vehicle distance measurement method using monocular vision and license plate

Estimating the distance between two vehicles is very important for transport safety. In order to keep the driving safety and avoid the traffic accidents, we proposed a rough distance measuring method using license plate to estimate the distance between two vehicles in this study. The method first uses checkerboard corner detection algorithm and the Tsai two-stage technique to calibrate the intrinsic parameters of the camera. Then the location of the license plate region in a single image are extracted based on the constrains of the color and the shape information of the license plate. Next, the corners is computed using the intersection points of parallel lines, which are extracted from the convex hull of the region of the license plate. Finally, the distance between two vehicles is estimated based on the intrinsic parameters of the camera, the image coordinates of the four corners and the size of the license plate. Experimental results of the static and moving cars show that our algorithm can estimate the distance between two vehicles, and the measurement error between our algorithm and a hand — held laser telemeter is within 0.8m. It means that our algorithm can be used as an efficient distance estimation method to keep the driving safety.