20 April 2016 Face recognition using 4-PSK joint transform correlation
Author Affiliations +
This paper presents an efficient phase-encoded and 4-phase shift keying (PSK)-based fringe-adjusted joint transform correlation (FJTC) technique for face recognition applications. The proposed technique uses phase encoding and a 4- channel phase shifting method on the reference image which can be pre-calculated without affecting the system processing speed. The 4-channel PSK step eliminates the unwanted zero-order term, autocorrelation among multiple similar input scene objects while yield enhanced cross-correlation output. For each channel, discrete wavelet decomposition preprocessing has been used to accommodate the impact of various 3D facial expressions, effects of noise, and illumination variations. The performance of the proposed technique has been tested using various image datasets such as Yale, and extended Yale B under different environments such as illumination variation and 3D changes in facial expressions. The test results show that the proposed technique yields significantly better performance when compared to existing JTC-based face recognition techniques.
© (2016) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Md. Moniruzzaman, Md. Moniruzzaman, Mohammad S. Alam, Mohammad S. Alam, } "Face recognition using 4-PSK joint transform correlation", Proc. SPIE 9845, Optical Pattern Recognition XXVII, 98450F (20 April 2016); doi: 10.1117/12.2223756; https://doi.org/10.1117/12.2223756


Back to Top