数据融合在车牌字符识别中的应用

将D-S证据理论引入车牌字符识别,结合神经网络容错能力强、能够自适应学习等优点,对待识别字符进行统计和结构等多方面的特征提取后,分别经过神经网络分类器得到相应的结果,并应用D-S证据理论对各种结果进行数据融合,从而实现了字符各方面特征的优势互补,进一步提高了字符识别的成功率。

The application of data fusion in the license plate character recognition

This paper introduces the D-S evidence theory to license plate character recognition. Combined with a series of advantages of the neural network of fault-tolerant capability and adaptive learning capability, statistics on the identified character and structure, and many other feature are extracted. Through the neural network classifier, corresponding results are gotten respectively, and data fusion is done by application of DS evidence theory results for a variety of results. Then, it achieves the characteristics of various aspects of the character strengths to further improve the success rate of character recognition.