Face Recognition based on Oriented Complex Wavelets and FFT

The Face is a physiological Biometric trait used in Biometric System. In this paper face recognition using oriented complex wavelets and Fast Fourier Transform (FROCF) is proposed. The five-level Dual Tree Complex Wavelet Transform(DTCWT) is applied on face images to get shift invariant and directional features along ±15o ,± 45o and ± 75o angular directions. The different pose, illumination and expression variations of face images are represented in frequency domain using Fast Fourier Transform(FFT) resulting in FFT features. Features of DTCWT and FFT are fused by arithmetic addition to get final features. Euclidean Distance classifier is applied to the features to recognize the genuine and imposter faces. The Performance analysis of proposed method is tested with ORL, JAFFE, L-SPACEK and CMU-PIE having different illumination and pose conditions. The Results shows that Recognition Rate of proposed FROCF is better compared to Existing Recognition Methods.