Face recognition algorithm based on discriminative dictionary learning and sparse representation

In order to overcome the defect that the face recognition (FR) rate is greatly reduced in the existing uncontrolled environments such as the change of illumination, occlusion, and posture, etc, Face recognition algorithm based on discriminative dictionary learning and regularized robust coding was proposed. In this proposed algorithm, the Gabor amplitude images of a face image are obtained via using Gabor filter at first, then we extract the uniform local binary histogram and use Fisher criterion to gain a new dictionary, finally the test image is classified as the existing class via sparse representation Coding. The experimental results obtained from Extended Yale B databases and AR databases show that the proposed algorithm has higher face recognition rate in the existing uncontrolled environments in comparison with K-SVD, LC-K-SVD, FDDL and so on.