Abstract: In order to effectively extract the global and local features to improve the performance of face recognition,a novel algorithm for face description and recognition based on LBP pyramid feature is proposed.By the algorithm,first the face image pyramid is constructed through a multiscale analysis.Then the LBP operator is applied to each level of the image pyramid to extract LBP feature maps which compose the proposed LBP pyramid.Finally,the feature maps in LBP pyramid are respectively divided into several regions from which the LBP feature distributions are extracted and concatenated into an enhanced feature vector to be used as the face descriptor.Experimental results on ORL and FERET face databases show that the proposed algorithm can be applied to achieve high face recognition rates.This work demonstrates that the LBP pyramid feature is highly discriminable with good performance in face feature expression and is robust to illumination,face expression and position variations.