IDENTIFIKASI PENGENALAN WAJAH UNTUK SISTEM PRESENSI MENGGUNAKAN METODE KNN (K-NEAREST NEIGHBOR)

Attendance is an activity that is so important and cannot be separated from a teaching and learning activity to calculate and see student attendance. In this Final Project, research on the automatic attendance system is carried out through facial recognition identification (face recognition) using a webcam as a system input, then the resulting image capture results from each image will be processed through feature extraction using the LBPH (Local Binary Pattern Histogram) method and classification with the KNN (K-Nearest Neighbor) method and the help of OpenCV library-based Python software. The research in this Final Project obtained an average accuracy value in facial recognition using LBPH (Local Binary Pattern Histogram) of 93.9%, with an average FAR value of 4.66% and an average FRR value of 1.33%. For the classification of KNN (K-Nearest Neighbor) using Euclidean Distance when k = 1 obtained an accuracy of 100% with a computation time of 34 ms, at the time of k = 3 an accuracy of 98% with a computation time of 37 ms was obtained and at the time of k = 5 an accuracy of 88% with a computation time of 42 ms.