Gabor-Zernike Features based Face Recognition Scheme

Abstract

A Facial recognition (FR) system in still images is an important application in computer vision and image processing. Extraction of invariant features is the core of FR systems. In this paper, we propose a novel and efficient facial image representation based on Gabor energy filters (GFs) and Complex Zernike moments (ZMs), where GFs is used for texture feature extraction and ZMs extracts shape features, Almost all existing methods use only magnitude component of the ZMs (respectively GFs) as features in recognition task. Recently it is well known that the phase component of moments (respectively Gabor Filters) also captures useful information for image representation, in other hand, a simple Genetic Algorithm (GA) is applied to select the moment features that better discriminate human faces and facial expressions, under several pose and illumination conditions. Next, the extracted feature vectors are projected onto a low-dimensional subspace using Random Projection (RP) technique. The (GF+ZM+RP) feature vectors are then applied to a powerful face support vector machine (SVM) classifier, employing the Gaussian radial basis function as kernel function (RBF kernel). Comprehensive performance evaluation of our proposed algorithm is based on Libor Spacek’s Facial Images Databases and Color FERET Database. It can be concluded from the experimental results that the performance of the proposed FR system outperforms other related approaches in terms of recognition rate.