Design and Implementation of an Intelligent & Real-Time Vehicle Access Control System

Abstract

This thesis explains how to combine learning software and image processing to construct a system that can extract a license plate number from a video of a car taken with a camera and recognize the driver's face. This project was offered by Icosnet which wants to improve the efficiency of its staff's workflow. We trained a convolutional neural network using different ways to detect objects, and we used the dataset we acquired to train it. After that, we develop a software solution that allows us to control automobile access by utilizing our database of personnel with access to the workplace parking lot. we make that by recognizing the license plate of their cars or by recognizing their faces. This solution gives us a high accuracy in a short execution time. Keywords: Image processing, License plate recognition, Access control, Face recognition, Convolutional neural network