基于主元分析的人脸特征提取MATLAB实现

Implementation of Facial Feature Extraction based on Principle Component Analysis by MATLAB

人脸的特征提取是人脸识别的关键技术,采用主元分析法进行特征提取是经典的方法之一,利用M at-lab进行人脸的特征提取能显著地提高计算效率。论述了利用主元分析和奇异值分解进行人脸特征提取的方法,并详细阐述其在M atlab中的实现过程,包括读取图像文件、计算均值脸、求特征值和特征向量,计算人脸特征参数。实现过程均给出了M atlab代码。实践证明利用M atlab进行主元分析提取特征是一种有效的方法。

Facial feature extraction is a key technique of face recognition. Feature extraction using principle component analysis is one of the classical ways. Facial feature extraction by Matlab can increase computational efficiency. The facial feature extraction based on principle component analysis and singular value decomposition is discussed in this paper. The implementation methods by Matlab, which include reading image files, computing mean face, computing eigenvalue and eigenvector, getting facial feature, also are discussed in detail. For all this Matlab source codes are presented. Practice shows that feature extraction using PCA by Matlab is an effective method. And this method is useful for the researcher in this field.