基于神经网络的硬币面额识别  

Coin Denomination Recognition Based on Neural Network

作　　者：毛玺[1] 郭光立[1] 

机构地区：[1]河南科技大学机电工程学院,河南洛阳471003

出　　处：《光电技术应用》2010年第2期54-56,共3页Electro-Optic Technology Application

基　　金：河南省自然科学基金(0411012500)

摘　　要：硬币面额的自动识别是自动售货机和硬币清分机具的关键技术.通过摄像头采集硬币的图像信号,利用硬币的外观进行面额识别.首先提取硬币图像的不变矩特征,然后利用神经网进行图像的分类识别,从而实现硬币面额的识别.实验表明该方法能得到满意的识别结果.The automatic recognition of coin denomination is a key technology for vending machines and coin sorting machines.The exterior appearance of the coin was used to recognition its denomination through capturing the image signal of the coin by using camera.Firstly,the invariant moments of coins image were extracted,then the neural network was used to classify the images for realizing the recognition of coin denomination.The experimental results show that the proposed method can achieve satisfying recognition results.

关 键 词：硬币面额 不变矩 BP神经网络 图像识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]