基于小波变换和支持向量机的人脸识别方法研究

Face Recognition based on the Wavelet Transformation and Support Vector Machine

考虑到人脸识别的非线性和小样本的特点,利用小波变换对人脸图像进行预处理,保留图像的低频段,有效地降低了图像维数并去除冗余噪声。采用奇异值分解的方法提取特征,然后利用支持向量机的方法设计人脸分类器,最后利用ORL人脸数据库进行验证,实验结果证明了该方法的有效性。

Considering the non-linear and small sample properties in face recognition, the wavelet transformation is used to reduce the dimension of the face picture. And the feature fetch is based on the singular value decomposition. The support vector machine is used to design a face classifier. The method is verified by using the ORL face database.