摘要
本文分别用主成分分析(PCA)、独立成分分析(ICA)以及线性鉴别分析(LDA)方法对图像进行特征抽取,采用支持向量机(SVM)算法进行人脸图像分类。通过在YALE人脸图像库上的实验结果验证表明,在多种特征抽取方法下的图像分类算法是有效的。
In this paper, a comprehensive performance comparison of face image extraction between three different feature extraction methods including PCA, LDA and ICA is made respectively. Moreover, the face image classification is also performed by using support vector machine （SVM） algorithms. Experimental results conducted on the YALE face image database demonstrate the effectiveness of the proposed method based on the different feature spaces.
关键词
特征提取
降维
支持向量机
人脸识别
feature extraction
reduce dimensionality
support vector machine
face recognition
