Abstract: A novel algorithm for face recognition is presented. The algorithm used Gabor wavelet transform and a new method for face feature extraction,calculates Locally Linear Embedding（LLE） for nonlinear dimensionality reduction of face data,and trained a nearest neighbor classifier for classification. The comparison among the proposed approach,PCA method,Gabor+PCA method and direct LLE method is carried out on ORL face database. The experimental results show that the proposed algorithm is superior to other methods.