Contrôle d’Accès par Vérification Parentale à base d’Images

Abstract

This thesis presents the development of an innovative access control system that verifies kinship based on facial characteristics from images. The main objective is to design a robust and accurate method to determine the relationship between individuals. This system has applications in residential security, access control to restricted areas, and identification within a family environment. The thesis utilizes convolutional neural networks (CNN) to classify similarities and differences among family members. The proposed method achieved an accuracy of 91.75% on the KinFaceW-II database.