Facial Biometric Identification in The Masked Face

Abstract

The COVID-19 pandemic has brought challenges in the field of biometrics to be able to carry out biometric identification on masked faces. Various studies on biometric identification on masked faces have been carried out and some have obtained promising results. This study aims to obtain a biometric identification method for masked faces using the JAFFE database dataset which has been manipulated into masked face images. The proposed method in this study can produce an accuracy value of 96%, which is promising enough to be applied in the biometric industry. The proposed method uses the face area segmentation technique and extraction of local binary pattern and histogram of oriented gradient features with the Support Vector Machine classification method.