基于Gabor小波特征的人脸识别算法研究

二维Gabor小波变换对图像边缘敏感,具有良好的方向选择性和尺度选择性,而对光照变化不敏感.本文提出了一种新型的Gabor特征人脸识别方法.该方法通过Gabor小波变换进行精确的人脸眼睛坐标定位,选取有利的人脸识别特征区域,通过构造新的Gabor小波特征,采用最近邻分类器对该特征进行分类.本文将该方法在YALE人脸库上进行了人脸识别实验,结果表明了该方法的有效性.

Study of Face Recognition Algorithm Based on Gabor Wavelet Features

Two-dimensional Gabor wavelet transformation of image edge-sensitiveness is good in direction selectivity and scale selectivity,but it is not sensitive to variation of light.A new method of Gabor features of face recognition is presented in this paper.The method,Gabor wavelet transformation,accurately positions the coordinates of the face,eyes and other favorable face recognition areas,and by constructing new Gabor features,classifies the features into the nearest category.Experimental results in YALE face ...