摘要
文章介绍了一种基于骨架法形态分析的连体字符图像识别方法,能较正确、容易地确定切分位置,使用修剪算法对各切分点进行处理,被分离的字符骨架可直接通过分类器而不必再作进一步处理。
A topographic approach recognizer is proposed for merged character images based on skeletonization. The breaking positions can be correctly and easily located.A splitting algorithm is then applied on each breaking position.The separated skeletons can be directly passed through a classifier without any further process.
出处
《计算机工程与应用》
CSCD
北大核心
2000年第2期65-66,161,共3页
Computer Engineering and Applications
基金
1997年度江苏省教委留学回国人员科研基金
关键词
连体字符图像
骨架法
字符识别
图像识别
Merged character images,Skeleton pattern,Splitting algorithm,Character recognition.