基于神经网络的数字识别研究

Digital Identification Study Based on Neural Network

本文介绍了神经网络的基本原理及BP网络，并根据数字字符识别的原理，设计了应用于识别的BP网络模型，最后采用VisualC＋＋6．0实现该BP网络。应用结果表明，该网络对数字字符识别具有很高的识别率，此网络的实现具有广阔的应用前景。

At first,this paper introduced the priciple of NN and a BP net,secondly,according to the principle of noised digital number recognition,it designed to the BP network model for applying to identify.At last,the BP network had been effectively carried out by using Visual C＋＋ 6.0 software.Practice shows that the network to identify digital character has a high recognition-rate.The author think that the realization of the network has wide application prospects.