Implementation of a Real-Time Student Attendance System in the Classroom Using YOLOv8 and Face Recognition

Implementasi Sistem Absensi Siswa Secara Real-Time di Kelas Menggunakan YOLOv8 dan Pengenalan Wajah

Abstract

Conventional attendance systems that rely on manual input often cause inefficiency and potential fraud. To overcome this, this study develops an automatic attendance system based on facial recognition using Face Recognition technology and YOLOv8. The system is implemented in Python using OpenCV, face_recognition, and PyQt5 as the graphical interface. The process starts with capturing and augmenting facial data, followed by storing face encodings in CSV format. During attendance sessions, the system detects faces in real-time through a camera, matches them with stored data, and records attendance based on arrival time with statuses such as “Detected,” “Late,” or “Not Detected.” Attendance data is automatically saved in Excel format. Testing at Universitas Muhammadiyah Sidoarjo shows the system can recognize over 95% of student faces accurately in less than 15 minutes after class starts. This proves the system is effective, efficient, and feasible to implement in both educational institutions and office environments.