Year: 2015
Type: conference-paper
Abstract: With the development of the computer, BOW model and SC model are applied to a large number of image classifications, and exhibit excellent performance, which become a hot topic in the field of computer vision.In the paper, we proposed a new framework of dictionary learning.the objective function based on sparse representation just consider the sparsity, while ignoring the spatial information of image and the correlation information of features, we apply spatial pyramid matching and add the discrimination fisher regularized penalty, by performing iterative optimization can get an excellent dictionary for representing image features, finally, we use max pooling and svm classifier for image classification.Experimental results in ORL and YALE face database show that the method has a high resolution.
Author: Yin Wei
Institution: Wuhan University of Science and Technology
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W2308945294