基于Gabor小波与共同向量的人脸识别方法  

Face recognition method based on Gabor wavelet and common vector

作　　者：火元莲[1] 齐永锋[2] 

机构地区：[1]西北师范大学物理与电子工程学院,兰州730070 [2]西北师范大学数学与信息科学学院,兰州730070

出　　处：《计算机应用》2009年第9期2383-2385,2388,共4页journal of Computer Applications

摘　　要：基于子空间的人脸识别方法易受光照、姿态和表情变化的影响,针对这一问题,提出一种基于Gabor滤波器与共同向量(CV)方法相结合的人脸识别方法。Gabor滤波器因其良好的方向与尺度选择性,能很好地提取图像局部特征,对光照、姿态、表情变化有一定的健壮性;共同向量方法是一种线性子空间分类方法,利用提取的同类样本的共同属性(共同分量)对测试样本进行分类,在训练样本较少的情况下能够取得较好的分类效果。通过在ORL与Yale数据库上的实验表明,提出的方法具有较好的识别效果。The performance of subspace-based face recognition methods is easily affected by variances of lighting, pose and expression. To overcome the limitation, a novel face recognition method was proposed which combined Gabor filter and Common Vector （CV） approach. Gabor filter could better extract the local features of images because of its selectivity on scale and orientation, and it is robust to variances of lighting, pose and expression. The common vector is a linear subspace classification method, which can obtain good classification results by extracting the common property of each class training samples. The experimental results show that the proposed method can obtain good recognition results on ORL and Yale databases.

关 键 词：GABOR小波 共同向量 人脸识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]