Traditional face recognition methods can’ t work well, if there are too many faces in the database. In order to solve the problem, this paper presents a face recognition method based on wavelet transform and principal component analy⁃sis. Experimental results show that the method can guarantee a high recognition rate, by greatly reducing the data dimension.%针对数据库中待识别人脸数目巨大,计算困难的问题,提出了一种基于小波变换和主成分分析的人脸识别方法。该方法首先对数据库中的人脸进行小波变换,然后对小波变换后的图像进行主成分分析提取特征,将特征送入BP神经网络进行训练、识别。经过对比实验,结果表明:该方法可以在保证较高识别率的前提下,大大降低数据维数,从而该方法可以应用于超大数据库的人脸识别。
展开▼