Facial Recognition Attendance System with Integrated Liveness Detection and Real-Time Reporting

Facial recognition technology can be an efficient and touch-free biometric technology for automating attendance in

academic and organizational environments. The paper presents a comprehensive Facial Recognition Attendance System with

Integrated Liveness Detection and Real-Time Reporting. The system is developed using Python and OpenCV library. The system

utilizes a webcam for facial images, Haar Cascade for face detection, and LBPH for face recognition. The attendance data is

stored in SQLite and displayed using a Flask-based web interface. The liveness detection component eliminates spoofing attacks

by images or videos by detecting eye blinks and movements through frame differencing and motion analysis. The system follows

a modular approach for data collection, model training, real-time face recognition along with liveness detection, and automatic

attendance recording. Experiments were conducted on a wide range of users in different poses and lighting conditions. The

facial recognition accuracy was 96.3%, liveness detection precision was 94.5%, and average face recognition time was 1.28

seconds. The system can run on normal desktop/laptop computers without a GPU and can be suitable for small and medium-scale

organizations. The facial recognition-based attendance system is a cost- effective and efficient solution that can replace existing

manual and RFID-based solutions. The system can record secure and touch-free automatic attendance data in real-time. The

system can be further extended to use deep learning techniques and can be integrated with other tools and technologies.