关键词: 车牌字符识别 多特征 相似性测度

Abstract:

A method for license plate recognition based on multiple features and weighted pattern similar：ty measurement is proposed for the embedded environmen t. Firstly, the multi-feature information including structure, profile and stroke is detected and the license plate code table is set up. Secondly,by we   ighted pattern similarity measurement,license plate recognition is realized. This method does not need samples and templates,costs little resources and  has high computing efficiency, and can meet the needs of resources and efficiency for license plate recognition in the embedded environment. The experim  ents show that the method has good robustness and high recognition rate.

Key words: license plate recognition;multiple features, similarity measurement