模板匹配联合特征法在车牌字符识别中的应用

Application of Template Matching and Feature Method in License Plate Characters Recognition

针对当前车牌字符识别中相似字符易出错的问题,提出了模板匹配联合特征的车牌字符识别方法。在算法设计中,首先,根据车牌字符特征构建存在相似字符,识别时容易出错的特征字符库;初次识别时,对所有车牌字符采用模板匹配算法;初次识别的结果如不属于特征字符库,识别结束;否则进行二次特征识别。通过对4000张不同环境下的车牌图片进行测试,表明本方法的识别率达到了99%,解决了车牌相似字符识别时易出错的问题。此外,本算法设计简单,识别速度快,适应性强,对车牌识别系统的普及起到了重要作用。

In order to solve the problem that the recognition of similar characters are easy to get wrong in the system of license plate recognition,the combination of template matching and character feature was put forward.At first,a feature character library was build based on the feature of license plate characters,which only included similar characters.All characters used the template matching algorithm when they were recognized for the first time.The recognition finished if the result was not included in the feature character library.Otherwise,feature recognition would be applied for them.The 4000 license plate images were tested that they had been taken in different environments.The result shows that the recognition rate is 99%,which solves the error recognition in similar characters.Furthermore,the algorithm is simple,fast,and highly adaptable,which will play an important role in the widespread use of the license plate recognition system.