An Automatic License Plate Character Recognition Algorithm Based on Template Matching
According to the problems of poor environmental adaptability,low recognition accuracy,and long recognition time in existing license plate recognition algorithms,an automatic license plate character recognition algorithm based on template matching is proposed.The collected license plate images are sub-jected to license plate localization,license plate preprocessing,and license plate segmentation to achieve accurate license plate recognition.The license plate localization is based on the color localization method,and effective pixels within the area are searched to ensure successful localization.In order to improve the denoising performance,discrete cosine transform and morphological processing techniques are used during the license plate preprocessing process.In order to improve segmentation accuracy,horizontal projection is used to obtain the upper and lower boundaries,and then vertical projection is used to obtain the character width for segmentation.Finally,the template matching method is used to recognize license plate charac-ters,which shortens the recognition time.The experiment shows that the algorithm has a short running time and a recognition accuracy of 100％.
万方数据
维普
