基于小波变换和线性判别分析的人脸识别方法

提出了将人脸图像的小波分解和线性判别分析结合以达到人脸识别的方法.首先对人脸图像作小波分解,并将分解后的低频系数进行线性判别分析进一步降低人脸特征向量的维数,最后利用最近邻分类器进行分类识别.实验表明,该方法的正确识别率高于传统的特征脸识别方法.

Face Recognition Based on Wavelet Transform and Linear Discriminant Analysis

This paper presents a face recognition algorithm that is based on wavelet transform and linear discriminant analysis.The face images are transformed into wavelet low frequency coefficients,which are processed by linear discriminant analysis for deducting the dimension of low frequency coefficients.At last,we use the nearest classification to recognize the face images.Experiment results show that higher recognition rate can be obtained than the traditional Eigenfaces method.