IoT-based Facial Recognition Door Lock with Push Button Combination

Abstract

A common issue in many households is the management of door keys, which can often be lost, damaged, or forgotten. This problem can be particularly frustrating when family members arrive home to find they cannot enter their residence. One potential solution to this problem is the implementation of a facial recognition system, which can be used by multiple family members to gain access to their home. The proposed system utilizes an intelligent minicomputer with a dual function, serving as both a Central Processing Unit (CPU) and a monitor screen display connected via an HDMI to VGA adapter. Additionally, the system incorporates a RealVNC Viewer (Virtual Network Computer) as a remote desktop application, accessible via a laptop or Android device. The primary aim of this research is to develop a facial recognition system that can unlock a house door by detecting and matching the user's face with stored data. The minicomputer, acting as a microcontroller, captures the user's face through a camera and compares it with the photo database. If a match is found, the system grants access and displays information on the Android device, indicating whether the user is known or unknown. This study employs a Research and Development (R&D) approach to enhance the effectiveness and security of home control systems using an internet network and Android devices. The system successfully detects facial objects within the camera's range, with the detection point area turning left, right, or focusing towards the front of the rectangular frame. The buzzer indicator sounds when the door lock is open. The system has been tested with multiple users, with local detection distance and over a well-validated network.