A Smart IoT Security System for Smart-Home Using Motion Detection and Facial Recognition
Akm Jahangir Alam Majumder, Joshua Aaron Izaguirre
Abstract
The Internet of Things (IoT) combines the idea of technology providing value to our daily lives. One major category in which technology benefits us is the concept of security and privacy. Smartphones can act as a security alert system as the smartphone is the most widely-used smart device. Recently, the use of Artificial Intelligence (AI)-integrated smart IoT devices has significantly increased as well. In this paper, we developed a smart IoT security system for the smart-home. In which, a Raspberry Pi acts as a security system with a No Infrared (NoIR) Pi Camera Module to record videos and capture images. Also, a Passive Infrared (PIR) Motion Sensor is used to detect motion. We propose to use motion sensor data and the images gathered from the NoIR Pi Camera Module to predict a security threat using a face recognition classification technique along with our developed algorithm. The system has the capability to notify the user in case of any emergency. The proposed system can detect any security threat with an accuracy of 95.5% and a precision of 91%.