基于整形特征和模糊识别的手写体汉字识别  

Manuscript Chinese Ideograph Recognition Based on Extracting of Integral Form Character and Fuzzy Recognition

作　　者：史绍强[1] 王英健[1] 唐贤瑛[1] 

机构地区：[1]长沙交通学院计算机工程系,湖南长沙410076

出　　处：《微机发展》2004年第1期114-116,共3页Microcomputer Development

摘　　要：针对以往手写体汉字识别方法的形近字准确识别率低、实现过程复杂和运算速度慢等不足,提出了一种比较简单和快捷的手写体汉字识别方法。该方法首先提取汉字整形特征,再按照笔顺进行编码,然后建立标准样本识别库,最后融合模糊模式识别方法以进行汉字识别。实验结果证明,该方法较好地解决了目前形近字的识别率低等问题,实现了以较快的速度和较高的准确率进行手写体汉字的在线识别。Aiming at some disadvantages of the previous manuscript Chinese ideograph recognition technique, such as low accurately recognizing rate, complex realization processing and slow operation speed,author brings forward an easy and shortcut scheme for the manuscript Chinese ideograph recognition.The scheme is that firstly extracting the integral form character of Chinese ideographs,and then coding with order of their strokes,in the following establishing a standard library of Chinese ideographs for recognizing,lastly coincided with fuzzy recognition algorithm to realize recognition process. Experimental results show that recent problem that the little efficiency of the similar ideographs'recognizing will be removed and the recognition will be completed at a high speed and with a good rightness.

关 键 词：手写体汉字识别 整形特征 模糊识别 人工智能 模式识别 图像处理 

分 类 号：TP391.43[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术]