基于神经网络的手写体汉字识别是将汉字点阵图形转换成电信号，然后输入给数字信号处理器或计算机进行 处理，依据一定的分类算法在众多汉字字符中找出和它相互匹配的汉字字符。本文阐述了手写体汉字识别实验系统的设计目 标，分析了手写体汉字的预处理及其原理，详细介绍了手写汉字的特征提取。

Handwritten Chinese characters recognition based on neural network is to convert Chinese characters dot matrix graphic to electrical signals, then input to the digital signal processor or computer for processing. On the basis of classification algorithm, it is recognized by matching with the Chinese characters. This paper expounds the design target of handwritten Chinese characters recognition experiment system, analyzes the pretreatment and the principle of handwritten Chinese characters, introduces the feature extraction of handwritten Chinese characters in detail