A Compressive Sensing Based Robust Face Recognition Method


Suparna Biswas

Author affiliation

Gurunanak Institute of Technology, 700114, India


In this paper an integrated face recognition method has been proposed to recognize the face images. This integrated framework combines the compressive sensing (CS) and Local Binary Pattern (LBP). The image is at first divided into small blocks and local binary pattern is generated corresponding to each block. Features are extracted from the LBP transformed image using blockwise histograms with variable no of bins. For classification we use compressive sensing based sparse representation classification (SRC). To study the robustness of our method, the classifier is studied on clean and occluded images. The experimental results give promising performance of the proposed face recognition method on JAFFE and ORL database


Face recognition; compressive sensing; local binary patterens