Enhancement of the Face Recognition Using a Modified Fourier-Gabor Filter

A modified Fourier-Gabor filter is used to enhance the classification rate of the face recognition. To verify the effectiveness of the proposed method, five well known methods are applied to four datasets; the methods are implemented without and with the suggested filter. The datasets consist of varying lighting conditions, different facial expressions, configuration, orientations and emotions. The experiments show that using the suggested Fourier-Gabor filter enhances the classification rates for all methods, all datasets and all training/testing percentage. The highest classification rates are obtained by using Fourier-Gabor filter with batch linear discriminant analysis (FGBatch-ILDA), where the average classification rate over the four datasets is 93.8, the next is 93.77 by using Fourier-Gabor filter with linear discriminant analysis (FG-LDA) and 90.85 by using FourierGabor filter with support vector machine (FG-SVM).