摘要
在车牌自动识别系统中,因自然因素或采样因素使得原本规则的印刷体字符产生畸变,给字符识别带来了很大困难.考虑到识别的实际应用及实时性要求,采用改进的模板匹配方法先将字符细化,再结合字符的全局特征,与模板的特征矢量进行匹配,误差最小的作为识别结果.实验结果表明,该算法具有较高的识别率及较好的抗噪性能,能够实现较好的车牌识别效果.
In automatic recognition system of license plate,due to natural or sampling factors the distortion of the regular printed character,brings a great difficulties for the character recognition.Considering the actual application and real-time requirements identification,the improved template matching method is used to the refine character,and match with the characteristic vector of the template after combining the inherent feature of the character.The minimum error is chosen as the recognition result.The experimental results show that the algorithm has high recognition rate and better antinoise performance,which can achieve good effect of license plate recognition.
出处
《山西师范大学学报（自然科学版）》
2014年第2期43-46,共4页
Journal of Shanxi Normal University(Natural Science Edition)
基金
高等学校科技创新基金资助项目(2013160)
山西省基础研究计划项目青年科技研究基金(2012021015-2)
关键词
数字图像处理
灰度化
模式识别
digital image processing
graying
pattern recognition
