Alignment of face images based on SIFT feature

In this paper, a method for the alignment of face images is proposed. We first extract SIFT features from the set of images, match features between each two images. Then we cluster the image according to the number of matching keypoints,in each category. We use the congealing algorithm to train and generate the model. The experiments show that the features are invariant to image scaling and rotation, and partially invariant to the change of illumination. This method has good performance in accuracy and in real-time.