基于主元分析的人脸识别方法研究

研究了利用主元分析进行人脸识别的方法。将人脸图像训练集进行主元分析,对得到的变换矩阵应用奇异值分解提取特征子空间,把训练图像和测试图像投影到子空间上,选择分类器进行人脸识别。实验表明,主元分析能很好地在子空间下提取出人脸图像的特征信息,从而实现人脸识别。

Research on Face Recognition Methods Based on Principal Component Analysis

A method for face recognition based on principal component analysis (PCA) is described. Transforming Face image training sets using PCA, Subspace of feature is extracted from the transformation matrix by singular value decomposition, the training images and test images is projected in the subspace, select classifier for face recognition. Experimental results show that PCA is well in the subspace of face images to extract feature information in order to achieve face recognition.