用自然结构分析法识别手写汉字

Recognizing Handwritten Chinese Characters Using the Natural Structural Analytical Approach

本文从汉字的自然层次结构出发,提出了手写汉字识别的自然结构分析法。在充分考查了国标集6763个汉字的基础上,建立了一个模拟实验系统。在该系统中,使用了扩充的有限自动机和启发式搜索等技术。该系统能识别带有少量连笔的手写汉字。实验结果表明:自然结构分析法是一种有效的手写汉字识别方法。

In this paper,a approach of the handwritten Chinese character recognition,called the Natural structural analytical approach, is proposed. A simulation system based on the approach is also described. The system includes a extended finite automata and heuristic search technique. And it can deal with some running-strokes. It is showed by the simulation experiment that the natural structural analytical approach is a effective approach for the recognition of handwritten Chinese character.