基于Gabor小波特征抽取和支持向量机的人脸识别

文章利用Gabor小波对位置误差、光线等因素具有强的鲁棒性的优点,将人脸图像在一定格点上取大小和方向不同的2D-Gabor小波变换,取变换系数幅值作为特征向量,送入支持向量机中进行分类。有效地结合了Gabor小波的特征抽取能力和支持向量机的分类能力,并对AT&T人脸库进行性别分类和人脸识别,得到了较高的识别率。

Face Recognition Based on Gabor Wavelet and Support Vector Machine

This article makes use of the advantages of2D-Gabor Wavelet filter,which are robust to the position error and illumination change.Gabor wavelet filter takes different amplitudes and directions.The response amplitudes of the image's2-D Gabor wavelet transformation at some grid points are fed into SVM(Support Vector Machine)network as feature vector for classification.The ability of2D-Gabor wavelet for feature extraction and the one of SVM for classifica-tion are combined,which is used for sex classification and face recognition on the face database of AT&T.And good recognition rates have been attained.