小波与多小波变换在支持向量机人脸识别中的研究

SVM Face Recognition Compare between Various Wavelet and Multi-Wavelet Transform

以ORL人脸库为基础,采用不同的小波和多小波,对人脸图像进行不同层次的小波和多小波分解,提取低频分量作为特征向量,利用支持向量机(SVM)进行人脸识别,通过实验具体给出了在采用不同的小波和多小波对人脸图像进行不同层数的小波分解时,支持向量机的识别率,该文对实际应用具有良好的指导价值和参考价值.

In this paper, the face images are decomposed by different wavelet and multi-wavelet at different level, and the low frequency components are translated into vectors, which will be input into the SVM. Experiment results base on ORL face database are shown in this paper, which will be useful in guiding realistic application.