Implementasi Otentikasi Multi-Faktor Menggunakan Liveness Detection dan Geolocation pada Sistem Absensi Online Berbasis Face Recognition

Atarza, M. Irpandi (2026) Implementasi Otentikasi Multi-Faktor Menggunakan Liveness Detection dan Geolocation pada Sistem Absensi Online Berbasis Face Recognition. Skripsi thesis, UIN Imam Bonjol Padang.

	Text (Cover - Daftar Isi) COVER - DAFTAR ISI.pdf - Published Version Download (1MB)
	Text (BAB I) BAB I.pdf - Published Version Download (269kB)
	Text (BAB III) BAB III.pdf - Published Version Download (390kB)
	Text (BAB V dan Daftar Pustaka) BAB V - DAFTAR PUSTAKA.pdf - Published Version Download (284kB)
	Text (Full Text) Full Text.pdf - Published Version Restricted to Repository staff only Download (2MB)

Abstract

With the rapid advancement of technology, many companies and institutions have shifted from manual attendance processes to more efficient digital attendance systems. However, digital attendance systems remain vulnerable to various security threats, particularly attendance fraud such as the use of fake photos, attendance submission from unauthorized locations, proxy attendance, and other forms of manipulation. This study aims to develop a web-based attendance system that implements Multi-Factor Authentication (MFA) by integrating face recognition, liveness detection, and geolocation. The research employed the Research and Development (R&D) method, which included the stages of requirements analysis, system design, implementation, and testing. The face recognition process was developed by extracting facial features using ResNet-34 and performing matching through the Euclidean Distance method. The liveness detection module was implemented to verify the presence of a real person through automatic eye-blink detection based on the Eye Aspect Ratio (EAR) algorithm. Furthermore, the geolocation module validated the user's attendance coordinates using the Haversine formula within a tolerance radius of 50 meters. The performance evaluation conducted on 20 respondents with 45 testing scenarios showed that the face recognition module achieved an accuracy rate of 78.33%. The liveness detection mechanism successfully detected eye blinks with an accuracy of 84.44%. In addition, the location validation module achieved an absolute accuracy of 100% in both accepting and rejecting attendance requests. The results indicate that the implementation of Multi-Factor Authentication combining face recognition, liveness detection, and geolocation effectively enhances the security of the attendance system. The limitations of one authentication factor can be compensated for by the others, resulting in a more reliable system compared to the use of a single authentication method.

Actions (login required)

View Item