License Plate Character Recognition Based On Combined Feature Extraction And BP Neural Network

In 2009,China became the biggest country of Auto pr oductionand consumption,and in the end of 2014, the amount of auto inventory reached 154 million, 117 million of them are ranch wagons.License Plate Rec ognation technology is the crucial procedure to assi st police handle the diverse communication affaires. Due to the complicate structure of chinese character s, chinese character Recognation is difficulty and e mphases of License Plate Recognition.Many domestical researchers have already issued many proposals to re cognaize plates. and currently the chinese character Recognation rate reach a very high level to the good quality license plates images.In this paper, we normalize the chinese character,co arse classificatethem and take improved BP neural ne tworks to design the Recognition device.The Recogna tion device will be with high Recognation rate and good anti interference capability.