Sistem Pendeteksi Wajah Bermasker Secara Real Time Menggunakan Metode CNN

Abstract

In this study, a masked face detection system using the Convolutional Neural Network (CNN) method was tested. System testing was carried out to determine the success of the CNN method on a masked face detection system in real time if the mouth and nose parts of the face are not covered or covered by a mask or by anything other than a mask. This research was divided into 3 stages, namely dataset training, face detection, and testing in real time with various facial positions. This system successfully detects whether a face is masked or not in the dataset model with an accuracy of 99%. In real time, this system has succeeded in detecting well at various positions of facial appearance and on faces that are covered by a mask or hand on the nose and mouth.