基于二维图像矩阵的ICA人脸识别

为了解决传统独立分量分析(ICA)在人脸识别过程中存在的高维小样本问题,同时为了提高识别效率,提出了一种基于二维图像矩阵的独立分量分析(ICA)特征提取方法.该方法将人脸图像矩阵作为训练样本,首先利用主分量分析(PCA)对训练样本进行去二阶相关和降维处理,然后对处理后的样本进行ICA特征提取,由于训练样本维数很小,因此它降低了传统ICA方法中高维小样本问题产生的识别错误率,同时减少了识别时间.在Yale人脸库和ORL人脸库上验证了该算法的有效性.

Independent component analysis for face recognition based on two dimension image matrix

To solve the problem that the number of available training samples is great less than that of the training vector in traditional independent component analysis (ICA) and to improve the efficiency of face recognition, an independent component analysis feature extraction technique based on two dimension image matrixes is proposed. The image matrix is taken as training sample directly, and principal component analysis (PC A) is used to reduce dimension and remove second-order correlation of training samples, then ICA is used to extract feature from the samples having been processed, since the dimensionality of the training sample is much smaller than that in traditional ICA, it can reduce the face recognition error caused by the dilemma in traditional ICA, and decrease the recognition time. Experiments on the Yale and ORL databases validate the effectiveness of the proposed method.