应用专门知识引导的联机自然手写汉字识别

研究一种识别联机自然手写汉字的方法，充分利用汉字知识，构造联机汉字的知识框架，在专门知识引导下识别自然手写汉字。汉字的多种静态构形规律和动态书写规律被用于联机汉字的知识表达和建模，输入汉字经过笔划元提取后，采用框架描述，与知识库中存储的汉字模型进行不精确匹配实现识别。对国标一级汉字进行了测试。在对书写次序、笔划数目、是否连笔等很少限制的情况下，能获得很高的识别率，证明该方法是行之有效的。

A Knowledge Based Algorithm for Online Cursive Chinese Character Recognition

This paper provides a knowledge guide online cursive Chinese character recognition system, where both static shape forming knowledge and dynamic writing are utilized. Every character is represented by a frame, which is used to store the various types of knowledge. After extracting of strokes, the input character is fed into an in exact matcher to perform recognition. Tests on GB 80 2312 Chinese character set demonstrate the proposed algorithm to be a efficient and robust algorithm with high correct rate , and with few restrict on stroke numbers, writing orders , and shape variations.