基于多重线性非参数化特征提取算法的人脸识别技术

摘要

针对人脸识别技术的特征提取方法存在不准确等问题，提出了基于张量空间模型的多重非参数化特征提取算法，给出了算法的基本原理，最后通过在ORL标准人脸数据库上的实验结果，验证了MNFA方法要明显优于MDA方法．%Due to the inaccuracy of feature extraction used in human face recognition, this paper proposes multilinear nonparametric feature extraction algorithm based on tensor spatial model, then gives the basic principle of the algorithm, and finally proves that MNFA method is superior to MDA method by experiment in ORL standard face image database.