Machine Learning-Driven Attendance System Using KNN and Image Processing

Abstract

Numerous time-consuming techniques are available to monitor an individual's presence, including biometric attendance tracking. Thus, image processing provides a more effective solution than biometrics or manual techniques. Particularly important for identification among physical characteristics is the face, with its plethora of distinctive features. This research presents a technique where a group image is initially captured, followed by individual face identification utilizing a face recognition module and the KNN (k-nearest neighbors) algorithm. Image capturing continues throughout the session, and attendance is recorded at the end once all individuals have been identified. Regular updates to the database have the potential to enhance accuracy over time. Individuals' attendance is recorded via Automatic Attendance Tracing (AAT) when the taken image matches the one stored in the database, indicating a match. This proposed approach reduces manual effort and streamlines the process of tracking student attendance, making it easier to register their presence.