Pembangunan Sistem Monitoring Kehadiran Mahasiswa Menggunakan Yolo Pendeteksi Obyek dan Pengenal Wajah Opencv

Abstrak

This research aims to address the issue of time-consuming and error-prone student attendance recording methods. The efficiency and accuracy problems in attendance recording motivate this study to develop a student attendance application based on computer vision technology using OpenCV and YOLO. The method starts with the collection and management of data, in the form of student facial photos, as the database. The facial recognition approach involves implementing the Simple Face Recognition library from OpenCV, while real-time facial detection is performed using the YOLO model. The integration of these two models is carried out in a server-based API using FastAPI. Furthermore, the development of a mobile application using Flutter enables users to capture and send student facial images to the API server and receive a response containing the list of detected students within the image. The objective of this research is to create an accurate, fast, and efficient student attendance recording solution through computer vision technology. The testing of the developed application shows promising results with a successful integration between OpenCV and YOLO technologies. This successful integration exhibits satisfactory performance, indicating substantial potential for this application to enhance attendance management efficiency within educational settings.