基于神经网络的手写体汉字识别实验系统

NEURAL NETWORK BASED HANDWRITING CHINESE CHARACTER RECOGNITING SYSTEM

本文针对BP网络的特点,讨论了手写体汉字特征的抽取,给出了BP网络曲算法,对于特定人手写体汉字大写“壹”～“拾”取得95％以上的识别率。

This paper presents a method for extracting and recogaiting the Chinese character, which is based on an BP network. Ninty five percent of the handwriting Chinese character can be correctly recognizied.