基于BP神经网络的人民币识别方法研究

随着信息化时代的到来,智能识别成为研究的热点,本文以人民币识别为研究对象,运用Matlab软件系统中所提供的神经网络工具箱,结合图像处理技术,实现对各种不同面值纸质版人民币的识别。本文主要针对六种不同面值的纸币,即一元,五元,十元,二十元,五十元,一百元的人民币进行了识别,首先通过样本集构建,图像预处理,特征提取,矩阵转置,样本与样本标签建立一一对应关系,生成相应的训练和测试矩阵,进而通过BP神经网络对数据矩阵进行分类训练,通过一定的训练次数后,该系统对人民币的识别率可达到98.33%以上,具有较高的参考价值。

Research on Identification Method of RMB Based on BP Neural Network

With the advent of the information age,intelligent identification has become a hot topic of research.This paper takes RMB identification as the research object,USES the neural network toolbox provided by Matlab software system and combines image processing technology to realize the identification of paper RMB of various denominations.This article mainly aims at six different denominations of notes,and the 1yuan,5yuan,10 yuan,20 yuan,50 yuan,100 yuan RMB for the identification,first by sample set build,image preprocessing,feature extraction,and matrix transpose,sample a one-to-one relationship with sample label,generate the corresponding training and test matrix,and then by the BP neural network to classify data matrix training,after a certain number of training,the system can achieve more than 98.33%recognition rate for the renminbi,which has high reference value.