论文部分内容阅读
本文以神经网络方法和字符识别研究为背景,提出了一种适用于汉字识别的并行神经网络(PNN)方法.它用一个称为控制网络的神经网络CN对汉字全集进行粗分类,用一组称为识别网络的神经网络RN对各粗类进行细分类,从而完成对汉字的识别.PNN方法与人类学习识字的过程相似,可以不断学习,最终完成对所有汉字的识别.PNN的各网络模块并行工作,具有极高的系统工作效率,并且其结构模块化,易于硬件电路实现.本文选取了120个汉字,用PNN神经网络模型进行学习和识别实验,还选取了30个汉字以及一些具有多种书写方法的汉字进行了追加学习实验.实验结果表明,PNN神经网络模型能够有效地应用于汉字识别研究.
In this paper, based on neural network and character recognition, this paper proposes a parallel neural network (PNN) method which is suitable for Chinese character recognition.It uses a neural network called control network CN to roughly classify the complete set of Chinese characters, The neural network RN, which is called recognition network, classifies each rough class to complete the recognition of Chinese characters.PNN is similar to the process of human learning and literacy, and can continuously learn and eventually recognize all Chinese characters.NP network modules Parallel work, with high system efficiency, and its modular structure, easy hardware circuit.This paper selects 120 Chinese characters, using PNN neural network model for learning and recognition experiments, also selected 30 Chinese characters and some have more Kinds of writing method of Chinese characters were carried out additional learning experiments.The experimental results show that the PNN neural network model can be effectively applied to the study of Chinese character recognition.