A multiview face recognition system based on eigenface method

Most of the face recognition methods are mainly focused on frontal face images. Identifying faces under different views still remain as a difficult task. In this paper, a new multiview face recognition system based on eigenface method is proposed. Principal component analysis is used to extract the features from the face image. Cr space image is chosen instead of gray level image to perform the pose estimation because it gives more information regarding the pose orientation. A pose estimation method based on eigenface method is proposed. The pose orientation of the face image is determined by estimating the likelihood with the mean faces of the database in eigenspace. Results show that Cr space image outperformed gray level image in pose estimation. Results also show that face recognition systems that take into account more views give better recognition results.