Analysis of Home Security System Design Based on 4 PIR Sensors Using Deep Learning Method

A home security system based on Passive Infrared (PIR) sensors has been developed and proposed by various researchers. PIR sensors are usually used to gather information about potentially hazardous object movements in the context of home security. Previous research has shown that the accuracy of movement detection depends on the number of sensors installed; the more sensors, the greater the accuracy. However, as the number of sensors increases, so does the detection time. Typically, systems with four sensors have an accuracy of about 90% and a detection time of 3 seconds. To address this issue, this research paper presents a solution using real-time deep learning models to shorten the detection time of suspicious objects. This research also evaluates the performance of various deep learning methods to determine the best result (in terms of object detection time) for home security. Rigorous experiments have been conducted to check the performance of the proposed system, and the results showed a detection accuracy of over 90%. In addition, we identify the potential of this system to enhance home security efficiency and effectiveness by combining high accuracy and faster detection times. Rigorous experiments yielded a detection accuracy exceeding 90%, instilling confidence in the system’s reliability. Thus, this research demonstrates that the implementation of real-time deep learning models in PIR sensor-based home security systems can improve response times to suspicious objects, enhance home protection, and provide an innovative solution to this challenge.