A License Plate Recognition Algorithm for Community Monitor

Abstract

In this paper, a novel license plate recognition algorithm is presented based on Hausdorff distance direct to community vehicle monitoring. This method utilizes colour and edge characters, meanwhile, combines some conventional image preprocessing methods to get several region of interest. In order to recognize the license plate, a well-organized digital models is prepared in initialization, and the Hausdorff distance is computed to recognize the vehicle characters according the threshold be defined before. The resulting of algorithm, which shows some tolerant of small location errors occurring in edge detection and has good robustness and efficiency.