Face recognition based on fractional discrete cosine transform

In this paper, we propose a robust and an accurate face recognition model which uses the feature extraction capabilities of fractional discrete cosine transform (FDCT). We apply FDCT on the face image database, and use transform coefficients as features. The proposed model is tested on publicly available standard AT&T face database to demonstrate the recognition accuracy. Extensive experimental comparison is provided with the DCT based face recognition algorithm to establish the superiority of the proposed model in terms of recognition accuracy.