Facial Recognition System Using Gabor Wavelet Transform, PCA Dimensionality Reduction, and SVM Multi-Class Classification

This program implements a facial recognition system that begins by processing face images using Gabor wavelet transformation to extract discriminative texture features. The Gabor transformation involves applying multi-scale and multi-orientation filters to capture facial characteristics at different frequencies and directions. Following feature extraction, Principal Component Analysis (PCA) is employed to reduce the dimensionality of the feature vectors while preserving the most significant variance in the data. This dimensionality reduction step optimizes computational efficiency and prevents overfitting. Finally, the system utilizes a Support Vector Machine (SVM) classifier from the libsvm toolbox to perform multi-class classification, where the SVM is configured with appropriate kernel parameters and multi-class strategies (one-vs-one or one-vs-all). The program is designed to work with the complete ORL face database and requires proper configuration of the libsvm toolbox before execution. The specific version used is libsvm-mat-2.89-3[FarutoUltimate3.0], which is the FarutoUltimate3.0 distribution of the libsvm toolbox for MATLAB.