A recognition method of Car License Plate Characters based on template matching using modified Hausdorff distance

In this paper, a novel recognition method is proposed aiming at the current difficulties existing in Car License Plate Characters recognition. Firstly, the method builds three Template Libraries apart including Chinese characters, English letters and Figure. The character to be recognized is preprocessed involving normalization, refinement and so on. And then, it is identified by using modified Hausdorff distance template matching method. The experiments show that the method is effective and feasible.