Affordable IoT Security: A Raspberry Pi System for Attendance and Intruder Detection
Karra Shivam Sharma, Ashish G. Hallur, Shreya Venkatesh, G. Sumathi, R. Sujatha, E. Konguvel
Abstract
This paper introduces an innovative and budget-friendly IoT-based facial recognition system tailored for attendance tracking and intruder detection, leveraging the power of Raspberry Pi. By utilizing the Local Binary Pattern Histogram (LBPH) algorithm, the system achieves remarkable accuracy and quick response times, outperforming traditional methods like Fisher-Face. It integrates real-time face detection with instant alerts to bolster security measures. The system not only automates attendance records, updating them in real time and sending them via email, but also identifies unrecognized individuals, capturing their images and notifying authorities promptly. Designed with cost efficiency in mind, it operates on low power and uses readily available components, such as an external camera and a PIR sensor for motion detection. This practical solution is particularly well-suited for use in hostels, schools, industrial facilities, and other environments requiring secure, contactless systems. Future plans for enhancement include incorporating advanced features like liveness detection, expanding the dataset for higher accuracy, and upgrading the hardware for improved portability and seamless IoT integration. These upgrades aim to position the system as a cutting-edge tool for modern surveillance needs.