Conception et réalisation d'un système de contrôle d'accès par biométrie et radiofréquence : cas de l'Enset d'Ebolowa.

Abstract

For several years in Cameroon, many schools (public and private) have been created in order to provide training for Cameroonian citizens capable of using their acquired knowledge and their skills on the job market. Thus as part of the creation of teacher training colleges in Cameroon, the ENSET of Ebolowa was born on November 24th, 2017 and is in its third year of existence and training of future teachers. We are probably seeing an increasing number of people who can get there and have access to the school building. From this observation, arise insecurity on campus and back and forth movements which are not always controlled. In view of the above observations, we design an on-board biometric and radio frequency access control system based on artificial neural networks (MLP) using the gradient back propagation algorithm. The objective is to automate the management of staff and visitor entry and exit within the ENSET campus of Ebolowa. As a result, we worked with a database of ten fingerprints and four RFID cards to manage identification and authentication. In order to make our system efficient and robust, we use PMC for the extraction of singular characteristics (core and delta) and the training of an imprint; RFID technology is based on near field magnetic coupling using a frequency of 13.56 MHz. Simulations are made using Matlab and Rsutdio software for the use of RNA in order to predict one of the classes of fingerprints that exist (Arc, Stretch Arc, Spire, Loop to the left and Loop to the right). The learning having been acquired, the integration of the MLP was made in the Arduino IDE for the operation of the system in the Arduino mega2560 microcontroller with information which is saved in a file in txt format to manage traceability.