The Human Face Recognition Based on the Improved Multi-scale LBP

Abstract

This paper put forward an operator for face recognition based on LBP algorithm and multi-scale fusion method. It uses an e-LBP operator to extract texture features of face images, then divides the LBP feature of images into non-overlapping feature areas, and computes the histogram statistics on them respectively. Finally the multi-scale LBP features are derived by connecting the LBP histogram of sequences in all areas, and the nearest neighbor classifier is used for face image classification and recognition. The experimental results shows that this algorithm is able to face a good image classification and recognition and can extract much richer global information.