中文摘要：

针对Gabor特征的稀疏表示分类方法中最小范数Z，稀疏求解精度的问题，提出了一种基于Gabor特征的全局稀疏表示的人脸识别算法．首先利用Gabor小波变换处理人脸图像得到ca-bor特征，建立超完备字典，然后在全局特征中引入向量总变差模型，并融合Gabor特征和全局特征，最后利用稀疏表示模型对融合后的特征进行优化．通过实验可以得出，这种新型人脸识别算法无论是对于图像的光照还是姿态和表情等多种变化因素都具备较强的鲁棒性．

英文摘要：

In view of the problem that the minimum norm l1 sparse precision of the Gabor feature sparse representation classification method, this paper proposes a global sparse representation face recognition algorithm that based on Gabor feature. First using Gabor wavelet transform processes face image to from Gabor feature, establishing a complete dictionary, then introduces the total variation model to the global feature vector, and fuse Gabor features and global features, finally using sparse representation model to optimize the characteristics of fusion. According experiments, whatever for the variation of light or facial expressions and so on, this new type of face recognition algorithm has strong robustness.