Image representation based PCA feature for image classification

For image representation methods of image classification, it is very important to represent the image well. In this paper, we propose a novel representation method for image classification, which can combine the advantage that the sparse representation can effectively use image information and the advantage that the PCA method can effectively eliminate the interference of irrelevant image information, and overcome the shortcomings of them. The proposed method firstly used the PCA method to obtain the first numbers of eigenvectors with the largest contribution rate for the samples of each subject as the training samples, and then these training samples are used to represent the test sample.