Authors: N. Rathika; P. Suresh; N. Sakthieswaran
Addresses: Anna University, Chennai, Tamil Nadu, India ' Department of Electronics and Communication, Vel Tech Dr. RR & SR R&D Institute of Science and Technology, Avadi, Chennai 62, Tamil Nadu, India ' Department of Civil Engineering, Anna University Regional Campus, Tirunelveli 627007, Tamil Nadu, India
Abstract: Face recognition system is an important area of human biometrics and a successful application of pattern recognition. Variable illumination conditions in face images are main obstacle in face recognition systems. To deal with this problem, this paper presents robust illumination-insensitive pre-processing, Zernike Moments-based feature extraction and L1-norm distance method. In pre-processing the input face image is transformed into an illumination-insensitive image by integral normalised gradient image (INGI) method. Features are extracted by Zernike Moments. Finally, face recognition task is performed by L1-norm distance measurement. Experimental tests on the CMU-PIE face database are performed. The proposed method shows 84% verification rate under different illuminations.
Keywords: CMU-PIE face database; face recognition; L1-norm; pre-processing; Zernike Moments.
International Journal of Biomedical Engineering and Technology, 2017 Vol.25 No.2/3/4, pp.267 - 281
Available online: 30 Oct 2017 *Full-text access for editors Access for subscribers Purchase this article Comment on this article