Deteksi Mata dan Alis Menggunakan Adaboost Classifier dan Haar Cascade

Since COVID-19 pandemic, using a mask is a daily necessity. Massive use of masks poses challenges to face recognition, surveillance cameras, age estimation, eye gaze tracking systems, and driver fatigue monitoring systems based on face detection. From these problems, a study is needed to be able to detect the eyes and eyebrows on faces that use masks. This study aims to implement eye and eyebrow detection using the Haar Cascade method. This research went through several processes, including pre-processing, integral image, training and testing using the Haar Cascade method. The results of the study show that the Haar Cascade method is successful in detecting eyes and eyebrows objects quite well with an accuracy rate of 95% on masked faces, 90% on bespectacled faces, 87% on bespectacled faces and tilted, 87% on bespectacled faces, tilted and closed eyes.