The International Journal of Innovative Research in Computer and Communication Engineering
Face recognition approaches faces many difficulties when there are variation in the face images due to lighting and other conditions. In this paper, an approach to recognize known faces based on Eigen vectors and a hybrid meta-heuristic feature selection algorithm is proposed. The eigenvectors which are covariance matrix of the face images together describes the difference between face images. Face recognition problem is viewed as a two dimensional recognition problem. Initially the face images are projected in to face space and using principal component analysis the eigenvectors with high eigenvalues are extracted to reduce the dimension of the feature vector.