"Design and Implementation of an Intelligent Authentication System based on Face Recognition"

Abstract

Today's institutions face major security challenges; as a result, they require a large number of specially trained personnel to achieve the desired level of security. As humans, these employees make mistakes that compromise the degree of security. As a solution to the aforementioned issue, a face recognition security system that can detect intruders to high-security areas and aid in the reduction of human error has been proposed. This system is comprised of two parts: software and hardware. The software consists of face detection and recognition algorithms based on CNN (Convolutional neural networks), while the hardware consists of a camera linked to an electric door lock. When a person wants to enter the zone in question, he must first authenticate using a touch screen, where a snapshot of his face is taken and sent to the recognition software to be analyzed and compared with a database of enrolled users; if the user is found, the door is unlocked (if not, it remains locked), and an access attempt is recorded in a database to be displayed on the control panel (website).