摘要
本文介绍了一种基于骨架法形态分析的粘连字符图象的分切方法，能较正确、容易地确定切分位置，经修剪算法对各切分点的处理。
A segmentation method of topographic approach 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
《中文信息学报》
CSCD
北大核心
1999年第2期40-45,共6页
Journal of Chinese Information Processing
基金
1997年度江苏省教委留学回国人员科研基金
关键词
字符识别
骨架法
形态分析
分切法
粘连字符图象
Character recognition Segmentation Merged character images Skeleton Splitting algorithm
