基于正交基非负矩阵与线性判别相结合的人脸识别

Research on Face Recognition Technology

非负矩阵方法是基于局部特征提取的算法。考虑到其没有类别鉴别能力，本文在融合非负矩阵与线性判勇基础上，提出一种改进的人脸识别方法，该方法将非负矩阵的基正交化，去除信息冗余，提高了计算速度。仿真实验纠证明了该方法的有效性。

Non-negative matrix method is based on local feature extraction algorithm.Considering the algorithm can＇t tell image＇s classification, this article combine NMF and LDA algorithm,and give an improved method.Through making vector of base matrix orthogonal,infor-mation redundancy is eliminated and the calculation speed is higher than before.The experiments results show that is effective.