Single Sample Face Recognition Based on Global Local Binary Pattern Feature Extraction

To improve the recognition rate of single sample per person (SSPP), in this paper we propose a novel single sample face recognition method based on global LBP feature extraction. We first calculate the LBP value of each pixel based on the whole image and obtain the corresponding LBP image. Then, we segment the LBP image into non-overlapping image blocks. For each image block, we take its statistical histograms as its global LBP feature. Finally, we use the nearest neighbor (NN) classifier for face classification. Experimental results on three widely used face databases, including AR, FERET and ORL databases, demonstrate the effectiveness and robustness of the proposed method.