基于神经网络的字母识别系统的设计与实现

Design and Implemen of Letter Recognition System Based on Neural Network

基于神经网络,采用M atlab 6.5和V isualC,设计一个字母识别系统。该系统通过对BM P图片的二值化处理,在VC环境下调用M atlab,并将把二值化后的数据进行网络训练,从而实现26个英文字母的识别。系统性能的测试表明,系统所训练的神经网络有很好的抗干扰能力。

A letter recognition system based on neural network is developed by using Matlab 6.5 and Visual C. In this system, the BMP image is binarized. The binarized data are used in the neural network for the recognition of 26 letters. The system test reveals that the neural network has a good anti-jamming.