Open AccessRANCANG BANGUN APLIKASI NEW NORMAL COVID-19 DETEKSI PENGGUNAAN MASKER MENGGUNAKAN HAAR CASCADE CLASSIFIER
Author(s) -
Ade Saputra,
Ahmadi Ahmadi,
Ariesta Lestari
Publication year - 2021
Publication title -
jurnal teknologi informasi (jurusan teknik informatika, fakultas teknik universitas palangka raya)/jurnal teknologi informasi
Language(s) - English
Resource type - Journals
eISSN - 2656-0321
pISSN - 1907-896X
DOI - 10.47111/jti.v15i2.3291
Subject(s) - computer science , haar like features , python (programming language) , artificial intelligence , computer vision , covid-19 , classifier (uml) , crowds , face detection , pattern recognition (psychology) , facial recognition system , operating system , computer security , infectious disease (medical specialty) , medicine , disease , pathology
During the COVID-19 pandemic, when in public places, it is required to apply the 4M health protocol, namely wearing masks, washing hands, maintaining distance, and avoiding crowds. In its implementation, there are officers who always maintain and remind people not to violate health protocols. Like remembering to wear a mask.The mask detection application is made as a computerized surveillance system that can store images of violations of the use of masks and provide warning sounds. Observations, discussions and literature studies are sources of data in this empirical research. Using Python as a programming language assisted with OpenCV for image processing. After passing through the 4 stages of Waterfall, namely Analysis, Design, Manufacturing and Development and Testing, an application is produced where the Raspberry Pi is a processing tool and images are captured from the camera module with a resolution of 1080x1024 px. This application can detect the use of masks with an accuracy of 90.5% using the Machine Learning Haar Cascade Classifier method. Where the condition of the face is a maximum of 30 degrees turned to the side and looked up