基于自然语言处理的车牌匹配技术

License Plate Matching Technology Based on Natural Language Processing

研究车牌的准确匹配识别问题。马路上车辆日益增多,车牌个性化的趋势也在蔓延,由于个性化车牌与常规车牌在字体、大小、颜色等特性存在差异,像素识别特征不明显,传统车牌识别方法无法有效解决个性化车牌与匹配库中车牌像素特征的差异,在库中根据车牌信息进行匹配识别存在识别准确率不高的问题。为了解决上述问题,提出了一种基于图像和自然语言处理相结合的车牌匹配识别方法。通过将车牌中的数字信息准确提取,转化成自然语言环境下的语义特征,由于语义特征不受像素因素影响,解决了像素特征不一致造成的漏识别问题,实现对车牌的准确识别。实验表明,利用改进算法实现了个性化车辆图像的正确识别,取得令人满意的效果。

Research the accurately matched license plate recognition. With the increasing of vehicles on roades, the trend of individualized license plates are spreading. Due to the individualized license plates are different form the conventional licenses in font, size, and color, it~ pixel recognition characteristics are not obvious. Traditional license plate identification methods can not effectively recognize the individualized license plates. In order to solve the above problems, this paper put forward a new method which combines the images and natural language processing algorithm to solve the matching of license plates. The digital information of license plates was accurately extracted and trans- formed into a semantic features in natural language environment, because the semantic features are not affected by pixel factors. Then missing identify problem caused by disagreement of pixel characteristics was solved to complete the accurately identification of license plates. Experiments show that the improved algorithm realizes the correct iden- tification of individualized vehicle license image and has a good effect.