The Face Recognition Algorithm Based on LBP Features and Collaborative Representation
Yuxia Fu · Computer and Information Technology · 2015
In order to improve the recognition rate of algorithms, a face recognition algorithm of LBP and collaborative named L- CRC is proposed in this paper. First the LBP features of the training images and testing images are extracted. The LBP features are blocked and dimensionality problem is resolved by PCA to obtain feature vector.The dictionary is composed of the train samples' feature vector, Then, the test samples' feature vector are collaboratively coded over dictionary. The regularized residuals is computed. Test samples is classified by the minimum residual.Extensive experiments on ORL and AR face database demonstrate that the L- CRC can obtain higher recognition rate,which have some reference values for practical application.