本文设计研究纸币识别器的主要功能模块,如图像特征采集的硬件与程序设计﹑纸币 识别等。设计了以中断方式实现的实时特征采集系统;在纸币识别部分采用神经网络模式识别系统中使用很广泛的PCA主分量分析方法, 提取矩作为特征,给出纸币识别器的BP网络分类器械的设计和实现。 关键字:uclinux;纸币识别系统;自动挂号机;MDB/ICP 协议 Abstract: The main modules, such as image sampling, bill identification, data sampling, in Bill Acceptors are designed and studied. A real-time data acquisition system which uses interrupt is designed. Size feature is chosen to identify the value of RMB paper used an analysed main heft means of PCA , selected as features,educed how to desgn and implement the BP neural network mode classifying device. Keywords: uclinux ; bill acceptor system; self-help registration machines; MDB/ICP protocol