摘要
提出了一种改进的模板匹配方法 .该方法是在传统的模板匹配方法的基础上 ,通过对字符特征区域的扩大和加强注意设计了一种改进的模板 ,以达到更有效的匹配结果 .将该方法应用于沪宁高速公路收费口处实拍的车牌图像库中 ,其平均识别率达到 97.1% .
In this paper, a modified templet matching method is proposed. Based on traditional templet matching method, the modified templet is desighed to achieve a more effective result by extending and strengthening the feature region of the characters. This method is applied to recognize characters in car license plate database of Hu-ning highway and the recognition rate is 97.1%. Experimental results show that the modified templet matching method has higher recognition rate and robustness.
出处
《小型微型计算机系统》
CSCD
北大核心
2003年第9期1670-1672,共3页
Journal of Chinese Computer Systems
基金
国家自然科学基金资助项目 (69875 0 0 9)
关键词
字符识别
模板匹配
图像处理
characters recognition
template matching
image processing
