Development of an IoT-based smart home security security system using face recognition

Mohammad Hassan, Mohamad Aiman Hakim (2022) Development of an IoT-based smart home security security system using face recognition. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Development of an IoT-based smart home security security system using face recognition.pdf - Submitted Version Download (479kB)
	Text (Full text) Development of an IoT-based smart home security security system using face recognition.pdf - Submitted Version Restricted to Registered users only Download (3MB)

Abstract

The home security system has become a requirement in every residence. Previously, the majority of doors were accessed with conventional methods such as keys, RFID cards, passwords, or patterns. However, incidents such as key loss have resulted in numerous unpleasant cases, including burglary and identity fraud. In order to solve this significant problem, a facial recognition system for home security was developed. The ESP32 module and the Internet of Things (IoT) were employed in this project to develop a highly effective door access control system. In addition to the on/off relay for door access, an ESP32 module was implemented to provide control. ESP32-Cam operates as the facial recognition module, it is employed as a vision sensor. In the meantime, IoT connectivity was mainly used to connect and transmit information that enables the end user to control door access using facial recognition live streaming. The door will normally open automatically if the system recognizes an authorized individual. Otherwise, the homeowner will be alerted and have the authority to determine whether or not to open the door through the virtual lock widget in the Blynk application. Up to 85% accuracy for the facial recognition system is the expected outcome of this project. At the end of this test, the researcher manages to get an accuracy of 86% for this face recognition development. This implementation is capable of achieving greater accuracy than the researcher's target accuracy of 85%. In conclusion, facial recognition and IoT security systems have been successfully implemented. By using face recognition, this project has effectively produced a high-quality security system.

Actions (login required)

View Item