In order to promote the accuracy and robustness of face recognition

a face description and recognition method based on multi-scale LBP feature is proposed in this paper. First

the original face image is decomposed into two levels by utilizing wavelet analysis. Then the LBP operator is respectively applied on two approximate images to extract LBP feature maps. The two maps are respectively divided into several regions from which the LBP feature distributions are extracted and concatenated into an enhanced feature vector which is the proposed multi-scale LBP feature. Finally

the multi-scale LBP feature is used as the face descriptor for classification and recognition. Experimental results on ORL face database show that the proposed method can achieve high face recognition rate which is up to 99%. This work demonstrates that the multi-scale LBP feature is highly discriminable with good performance in face feature expression and is robust to face expression and position variations.