基于连通域的模板匹配法用于字符识别的算法研究  

Algorithm Research of Templet Matching Method Based on Connected Area Applied to Character Recognition

作　　者：郭晓松[1] 孔祥玉[1] 杨必武[1] 

机构地区：[1]第二炮兵工程学院,西安710025

出　　处：《计算机工程与应用》2000年第12期46-47,共2页Computer Engineering and Applications

摘　　要：文章介绍了一种用于字符识别的基于连通域的模板匹配算法.该算法是在传统的模板匹配法的基础上,运用了重心重叠等处理方法,并且对字符块进行连通域处理后再实施匹配,最后利用最大相似法得出识别结果.该算法在光学仪器读数图象的动态识别中取得了良好的识别效果.In the paper,an templet matching algorithm based on connected area is developed to recognize character.Based on traditional templet matching method,the algorithm uses such methods as barycenter superposition,applies connected area disposal to the recognized character and then matches,gets result by the maximal similitude degree in the end．The algorithm gaines excellent recognition result in the dynamic recognition of reading image of optical instrument．

关 键 词：字符识别 模板匹配 连通域处理 算法 图象识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]