A dynamic-template based method of rapid locating vehicle license plates

A new method of rapid locating vehicle license plates is studied, which is based on a kind of proposed dynamic-template. The aim of this method is to improve license plate recognition system to have a better robustness, a quicker executive speed and a better adaptation. In this study, the proposed technique integrates a new dynamic template based on the detecting of the position of candidate characters in license plate into license plate recognition system. A robust and rapid algorithm to obtain binary image from color license plate is first proposed, which is a region based approach and can overcome many common image flaws. Based on binary image converted, a rapid component-labeling algorithm using contour tracing technique is used to get all connected components and their bounding boxes. According to the features of license plate, a cleaning algorithm. is proposed to remove those obvious non-character connected-component candidates. Then a matching template, which is generated dynamically according to these detected character connected-component candidates and the features of specific license plate, is presented to search all character connected-components in a license plate and locate them. So the positions of all characters and their contours can be known, which are very helpful to the following character recognition. The implementation of this method is also studied. Test results show that this method is very efficient.