Implementasi Sistem Presensi Pegawai Berbasis Pengenalan Wajah Dengan Model Prototype Pada Percetakan Cresna Na

Abstract

Managing employee attendance accurately is a core part of human resource administration, since it directly shapes how efficiently an organisation runs and how trustworthy its records are. At Cresna NA Printing, attendance is still logged by hand through paper sign-in sheets, a practice that leaves room for input mistakes, slow recap processes, and even falsified entries. To address this, the present study builds a facial-recognition-based attendance system for Cresna NA Printing, developed through the Prototype model. Development moved through several stages: gathering user requirements, designing the system, building successive prototypes, evaluating and refining them, then implementing and testing the final product. The resulting application supports face enrollment, identity verification, check-in/check-out logging, employee-data administration, and role-based attendance reporting. To confirm a face match, the system checks that the Euclidean Distance stays at or below 0.45 and that confidence reaches at least 55%. Three testing approaches, White Box, Grey Box, and Black Box, were carried out, yielding 216, 138, and 189 passing scenarios respectively, with every scenario behaving as specified. Quality was further assessed against the ISO/IEC 25010 standard, looking specifically at Functional Suitability, Usability, and Security, which scored 92.05%, 92.44%, and 92.77% respectively, all falling in the excellent category. These results indicate that the facial-recognition attendance system is well suited to raising the efficiency, accuracy, and security of attendance management at Cresna NA Printing.