摘要
本文对现存三种汉字的计算机表示和两种传统的汉字结构分析方法进行评述。应用抗扑和几何的基本原理,分析了汉字结构及其制约关系,确定了四类组成汉字的基本关系。在此基础上实现了汉字原型。最后给出了把汉字原型应用于手写汉字识别的实例。
This paper discusses three kinds of Chinese character representations in a computer and two traditional analytical methods of Chinese character structure. By applying fundamental principles of topology and geometry to Chinese character structure and their constrained relations, four catalogues which contain basic relations to Chinese character structures are defined and Chinese character prototypes are constructed. Finally, an experiment is given to show how to use the prototypes in recognizing handwritten Chinese characters.
出处
《计算机研究与发展》
EI
CSCD
北大核心
1996年第8期634-640,共7页
Journal of Computer Research and Development
关键词
汉字结构
汉字原型
汉字识别
信息处理
Chinese character structure, Chinese character prototype, handwritten Chinese character recognition.