摘要
提出了一种基于笔划的一级分类、笔划特征二级分类的新方法来实现联机手写体汉字的识别。该方法对笔划变形的容忍度大、计算简单。通过两级分类,对国标二级汉字的识别率为98%以上,不要求笔顺,识别时间短。
Based on the first classifying of stroke and the second class of stroke characteristics a new method of completing online handwriting Chinese character recognition is presented.There are several advantages with online handwriting chinese character recognition such as simplicity,high degree of tolerance to stroke variance.and recognition of up to 98%.
出处
《重庆大学学报(自然科学版)》
EI
CAS
CSCD
1998年第3期79-82,共4页
Journal of Chongqing University
关键词
汉字识别
笔划分类
笔划特征
模式识别
手写体
Chinese character recognition
stroke classifying
stroke characteristics / stroke combination
distance Rule