Pendeteksian Masker dan Klasifikasi Masker Menggunakan Metode Region-Based Neural Network

Main Article Content

Syawal Ludin
Chairisni Lubis
Teny Handhayani

Abstract

At this time the world is experiencing a pandemic, the virus is COVID-19 and to prevent a very fast spread, there are many ways to spread the virus, starting from touching, one of which is through saliva when sneezing or talking, therefore all people around the world The world is given rules for washing hands, social distancing and wearing mas.


However, it is very unfortunate that there are still many who do not comply with the rules made. Due to this, the mask detection system exists to facilitate community monitoring to be more obedient to the regulations that have been made.


In the proposed system the Region-based Convolutional Neural Network (RCNN) is used to classify images which consist of three classes including medical masks, non-medical masks and not using masks. Later the system will detect people in one image. With the Region-based Convolutional Neural Network (RCNN) method, 2 experiments were carried out on 30 epochs with 2 different layers and the first layer got 86% accuracy and 74% accuracy validation and the second layer got 80% accuracy and validation by 79%. With the level of accuracy obtained, it is hoped that it can help the government in slowing down the rate of increase in the number of COVID-19 and also that the community can be more obedient to the rules that have been applied.

Article Details

Section
Articles

References

Sunario Megawan dan Wulan Sri Lestari, “Deteksi Spoofing Wajah Menggunakan Faster R-CNN dengan Arsitektur Resnet50 pada Video”, Jurnal Nasional Teknik Elektro dan Teknologi Informasi, Vol. 9, Nomor 3, (Agustus, 2020) h. 261-267.

Muhammad Abdul Rahman Irham Harfi dan Dedi Ari Prasetya. “Prototipe pendeteksi masker pada ruangan wajib masker untuk kendali pintu otomatis berbasis deep learning sebagai pencegahan penularan covid-19”, Simposium Nasional RAPI, XIX Tahun 2020 FT UMS, Tanggal Akses 27 juni 2022

LeCun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature Volume 521, 436–444 (2015). https://doi.org/10.1038/na ture14539.