Performance Evaluation of Face Recognition System using various Distance Classifiers

Abstract

Face recognition applications are gaining popularity day by day. Feature extraction, selection, and recognition are the three main steps of face recognition system. Recognition is done using classifiers as these play a vital role in making the system recognize the faces accurately to the extent possible. This paper evaluates the performance of the system using four different distance classifiers over ORL databases. DCT (Discrete Cosine Transform)-PCA (Principal Component Analysis) and LDA (Linear Discriminate Analysis) methods followed by Cuckoo Search algorithm have been used for extraction and selection of important features respectively. The results demonstrate the efficiency and efficacy of the face recognition system upon using Euclidean distance classifier.