摘要
随着交通事业的迅速发展,人工管理方式已不能满足实际的需求,针对这一问题,提出了汽车牌照字符识别系统设计,采用了数字图像处理技术、车牌定位技术、车牌字符分割技术、BP神经网络识别技术来解决车牌字符的识别问题,研究了图像预处理、边缘提取、车牌定位、字符分割、字符识别5部分内容,并给出了BP神经网络实现车牌字符的识别,实验表明,此方法能实现良好的定位精确度和较高的识别率.
With the rapid development of transport, labor management can not meet the actual demand. For this problem, a car character recognition system design is proposed. In this article, the license plate character recognition problem is solved, using digital image processing technology, license plate localization technology, license plate character segmentation technology and BP neural network recognition technology and so on. The researched main content is divided into five large parts： image preprocessing, edge detection, license plate localization, character segmentation, character recognition. The license plate character recognition is implemented by means of BP neural network. The experiments show that this method can realize good positioning accuracy and high recognition rate.
出处
《哈尔滨理工大学学报》
CAS
2012年第1期90-95,共6页
Journal of Harbin University of Science and Technology
基金
哈尔滨市创新人才研究专项资金(2011RFQXC006)
黑龙江省研究生创新科研资金(YJSCX2009-041HLJ)
关键词
车牌定位
字符分割
字符识别
BP神经网络
license plate location
character segmentation
character recognition
BP neural network
