RECONOCIMIENTO DE ROSTROS UTILIZANDO ANÁLISIS DE COMPONENTES PRINCIPALES Y ECUALIZACIÓN DE HISTOGRAMA

Abstract

Biometric systems are of great importance because of its multiple applications, ranging from business applications to security applications, which requires high efficiency. The face recognition is a research area with many applications since the ‘80s. Face Recognition is probably the easiest biometric method to understand because it identifies people by their faces, as human beings do. This work proposes the histogram equalization as a phase in the preprocessing which will attempt to bring the histograms of the faces to uniformity, this procedure is carried out in different ways and also use the Fast Fourier Transform to obtain the phase faces and then extract their features using Principal Component Analysis. Once the feature vectors are obtained from each person the Support Vector Machine will be trained to be used as classifier. This thesis has a theoretical framework that explains the main features of the methods used in the proposed system. The results of the system are presented in Chapter IV.