Multiple facial instance for face recognition based on SIFT features
Yunyi Wang, Chunqing Huang, Xiaobin Qiu · 2009
The SIFT features performs quite well on object identification. The features are invariant to image scale, rotation and robust to the noise and illumination. However, the face recognition based on SIFT features have rarely investigated systematically. This paper proposes a simple approach to solve face recognition problems based on SIFT features. It made use of three frontal facial image per person as training image and applied matching features threshold to accept or discard face image, hence there more than one image is available to choose the image of maximum matching features of these image for query image. The experimental results demonstrate the efficacy of the proposed.