基于多层BP神经网络的扑克牌识别

The recognition of the Play Card imagesby the BP neural network

实现了多层BP神经网络识别扑克牌图像的方法,首先使用C语言来编写BP神经网络的算法以及用VC++来实现对扑克牌图像的分析,然后再使用大量已知花色和牌点的扑克牌输入,作网络训练,训练完成后,利用网络识别样本扑克牌和带噪声的扑克牌。结果表明,该方法能够较准确地识别GIF扑克牌图像。

This paper introduce the method to recognize the NN by the BP（Back-Propagation Network） neural networko In this paper it is more important that the arithmetic of the BP neural network and the analysis of the color of the playing cards by the VC＋＋. The recognition of the images by the BP neural network is the key point in the paper and it shows that it is one of the good way to recognize the cards.