基于2DPCA和流形学习LPP算法的人脸特征提取应用  

Face Feature Extraction Based on 2DPCA and LPP Manifold Learning Algorithm

作　　者：李国芳[1] LI Guo-fang （College of Big Data and Information, Guizhou University,Guiyang 550025,China）

机构地区：[1]贵州大学大数据与信息工程学院,贵州贵阳550025

出　　处：《电脑知识与技术》2014年第11期7438-7441,共4页Computer Knowledge and Technology

摘　　要：人脸图像的特征提取是人脸识别系统中最关键同时也是难题之一。流形学习算法是近些年的人脸识别和语音识别两个领域应用较多的非线性降维方法。通过对人脸识别系统的研究,现提出一种全新的基于2DPCA（Two-Dimentional PCA）和流形学习LPP（Locality Preserving Projections）算法的特征提取方法,可为今后深入研究人脸识别技术提供较好的参考。仿真实验表明,该算法与传统特征提取PCA、LDA算法相比,可以取得更好的识别率。Face-image feature extraction is one of the key technologies and problems in face recognition systems. Manifold learning algorithm, as a non-linear dimension reduction method, has been used in facial recognition field and speech recognition field in recent years. A new feature extraction based on 2DPCA（Two-Dimentional PCA） and LPP（Locality Preserving Projections） algorithm of the manifold learning is proposed through systematic study of facial recognition system. And it may provide a good reference for further study of facial recognition technology. The simulation experiment shows that this algorithm has better recognition rate as compared with PCA, LDA algorithms of traditional feature extraction.

关 键 词：流形学习 人脸识别 特征提取 2DPCA算法 LPP算法 

分 类 号：TP18[自动化与计算机技术—控制科学与工程;自动化与计算机技术—控制理论与控制工程]