对联机手写汉字识别技术的研究——实现笔画识别  

Research of on-line Handwritten Chinese Character Recognition: Implement Stroke Recognition

作　　者：俞庆英[1] 吴建国[1] 

机构地区：[1]安徽大学计算机科学与工程系,安徽合肥230039

出　　处：《微机发展》2004年第10期68-69,72,共3页Microcomputer Development

摘　　要：联机手写汉字识别(OLCCR),是指用笔在图形输入板上写字,人一边写,机器一边认,是一种方便的汉字识别手段。在各种自动识别输入的方法中,OLCCR是能够代替或部分代替人工编码输入的惟一可能的方法。识别中主要是两方面的问题:建立汉字识别库和手写板上笔画轨迹的识别。文中就第二方面即手写笔画识别的问题进行了全面的研究,采用笔画基元帮助分析笔画轨迹,并用可视化编程工具VisualC++6.0实现了基于这种方法的笔画识别过程。On-line handwritten Chinese character recognition(OLCCR) refers to that the computer can recognize Chinese characters while people are writing on a data tablet.It's a convenient means by which people can input Chinese characters.In the process of recognition,there are two primary problems needed to be solved.One is establishing a library of Chinese characters,the other is recognizing the track of strokes on the handwritten board.This paper mainly deals with the latter,that is to say,it is written about the strokes recognition.And in the midst of research,a kind of visual program tool:Visual C++6.0 is utilized.

关 键 词：联机手写汉字识别 笔画 基元 模式识别 

分 类 号：TP391.43[自动化与计算机技术—计算机应用技术]