License Plate Binarization Method Based on Fuzzy Classification

Many conditions can make it difficult to binarize license-plate image such as nonuniform illuminated, degraded license plates and so on. Almost all the binarization method select right threshold from gray-level image without considering the color feature of license-plate image. According to the color and brightness characteristics, this paper proposes a license plate binarization method based on the algorithm of fuzzy c-means cluster using multi-layer classification of the plate. Experimental results show that this algorithm is robust in dealing with nonuniform illuminated, degraded license plates. With the proposed method, the recognition rate of the system can be improved significantly.