Smart Attendance and Student Monitoring System

DOI

https://www.doi.org/10.59256/ijsreat.20260606005

Published In

Volume 6, Issue 6 (2026)

Published Online

November-December 2026

Pages

25-34

Article Type

Original Article

Check for updates

Traditional classroom attendance systems rely on manual roll calls, signatures, RFID cards, or other dedicated identification mechanisms that require additional effort and may be vulnerable to proxy attendance. In addition, conventional attendance systems generally provide little information about classroom activities after attendance is recorded. This paper presents a Smart Attendance and Student Monitoring System that integrates computer vision, facial recognition, object detection, facial landmark analysis, voice-based alerting, database management, and a web-based dashboard into a unified classroom monitoring platform. The proposed system uses a camera as the primary sensing device. Student faces are detected using the YuNet face detector and recognized using the SFace face-recognition model available through OpenCV. Recognized students are automatically recorded in a SQLite-based attendance database. The system further integrates YOLOv8 object detection for identifying mobile-phone usage and MediaPipe Face Mesh for facial landmark extraction and drowsiness analysis based on the Eye Aspect Ratio (EAR). Temporal confirmation mechanisms are used to reduce false alerts caused by individual frames. Confirmed phone-use and drowsiness events additionally trigger an offline text-to-speech voice alert generated using pyttsx3, providing immediate in-class feedback without requiring the teacher to watch the screen. Attendance, phone-use alerts, drowsiness events, session presence, and derived engagement information are then stored in a centralized database. A Flask-based REST backend provides access to the stored information, while a web dashboard displays attendance summaries, alerts, engagement trends, student information, absentees, and session reports. Automated PDF report generation and optional email delivery further extend the system’s administrative functionality. The proposed architecture provides a practical foundation for intelligent classroom management using existing camera infrastructure without requiring specialized attendance hardware.