Electron: Jurnal Ilmiah Teknik Elektro
Vol 2 No 2 (2021): Jurnal Electron, November 2021

Sistem Pendeteksi Penggunaan Masker dengan Metode Convolutional Neural Network pada Sistem Portal Otomatis

Alwi Fran Fahlifi (Institut Teknologi Sumatera)
Heriansyah (Institut Teknologi Bandung)
Afit Miranto (Universitas Lampung)



Article Info

Publish Date
30 Nov 2021

Abstract

The very fast spreading process of COVID-19 has made this virus a pandemic in various countries. To reduce the spread of the COVID-19 virus, it is mandatory for everyone to follow health protocol rules such as social distancing and wearing masks. The health protocol examination is carried out by special personnel before entering the mandatory area to use a mask, which of course this examination will require more energy and cannot be done every time. In this study, a tool was made that could detect health protocols which would later reduce the workload of special workers. This tool can detect the use of masks on a person, which is made using the MobileNetV2 architecture and the Convolutional Neural Network (CNN) method that classifies people as not wearing masks and wearing masks. This tool uses the Raspberry Pi as a mini computer which is the main brain by adding a camera sensor to detect someone using a mask in real-time, RGB LEDs as a marker of whether the mask is detected or not, and the LCD as a display when the system is running. The effective distance that can detect the use of masks is as far as 30-200 cm.

Copyrights © 2021






Journal Info

Abbrev

electronubb

Publisher

Subject

Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering Energy

Description

E-journal of the Department of Electrical Engineering, Faculty of Engineering, University of Bangka Belitung, is a media for publication and information for scientific papers, undergraduate thesis, research, planning and design concepts, and analysis from students, professors, or any authors ...