摘要
介绍了一种识别车牌中汉字字符的方法 ,它包含了车牌的定位 ,预处理 ,字符识别等几个重要环节 .该识别方法分为 2级 ,粗分类是依据文字的四周面积编码来分类的 ;单个汉字的识别 ,是依据文字的全局笔划方向特征来识别的 ,由于此识别方法简单 ,识别速度较快 ,故识别效果理想 .
This paper introduces a way to recognize Hanzi in LP images. The identification system design can be divided in several stages: the license plate location in an image, the per processing, the character recognition and so on .In the paper, character recognition is divided in two steps: general classification and matching. In the first step, the surrounding area code is used to classify the character into one subclass. In the second step,we offer G DCD to match the character against a set of predefired templates. The method of recognition and caculation is so simple, that our character recognition is very fast and efficient.
出处
《中南民族大学学报（自然科学版）》
CAS
2002年第3期33-36,共4页
Journal of South-Central University for Nationalities：Natural Science Edition
