人民币面值特征参数提取及面值识别方法的研究

Feature Parameter Extraction and Face Value Recognition of RMB

本文在获得人民币纸币图像的基础上,先对图像预处理,然后提取特征参数,进而进行面值识别。在特征参数提取过程中,通过观察,使用基于小波变换的方法提取纸币面值区域的图像纹理特征和颜色特征。在面值识别过程中采用的BP神经网络算法,具有很强的鲁棒性,较好地实现了人民币面值的识别。

In this paper,on the basis of obtaining the image of RMB paper currency,the image is preprocessed,and then the characteristic parameters are extracted,and then the face value is recognized.In the process of feature parameter extraction,the wavelet transform based method is used to extract the image texture and color features of the banknote area through observation.In the face value recognition process,the BP neural network algorithm has a strong robustness,and realizes the face value recognition of RMB well.