Monitoring Pemakaian Masker Berbasis Video Menggunakan Metode Faster R-CNN

Romadhon, Muhammad Syahrul (2022) Monitoring Pemakaian Masker Berbasis Video Menggunakan Metode Faster R-CNN. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06111740000078-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2024. Download (11MB) | Request a copy

Abstract

Coronavirus Disease 2019 (COVID-19) menjadi masalah kesehatan utama yang menyebabkan penyakit pernapasan akut pada manusia. Tidak adanya kekebalan tubuh terhadap COVID-19 meningkatkan kerentanan terpapar virus, serta belum adanya approved drug mengakibatkan upaya dalam mengendalikan penularan wabah COVID-19 sangat bergantung pada intervensi non-pharmaceutical seperti upaya pencegahan secara perorangan, sebagai contoh penerapan protokol kesehatan. Pada 2020 lalu, WHO mengumumkan anjuran mengenai penerapan protokol kesehatan, salah satunya penggunaan masker guna mencegah penularan wabah COVID-19. Namun anjuran tersebut akan percuma, bila tidak ada kesadaran tiap individu masyarakat untuk memakai masker sesuai dengan prosedur yang ada. Dari permasalahan tersebut, dibutuhkan monitoring untuk menjamin pemakaian masker diterapkan sesuai dengan prosedur yang ada. Monitoring dapat dilakukan secara manual, tetapi membutuhkan biaya yang mahal dan sumber daya yang tidak sedikit. Oleh karena itu, dalam penelitian ini dirancang sistem monitoring otomatis dengan melakukan deteksi terhadap pengguna masker. Dalam penelitian ini, terdapat tiga jenis kelas (label) penggunaan masker, yaitu Pemakaian Masker Benar (Mask), Pemakaian Masker Salah (ImproperlyMask), dan Tidak Memakai Masker (NoMask). Proses pendeteksian penggunaan masker dalam penelitian ini dilakukan pada tahap testing, dengan langkah-langkah yaitu input video, akuisisi video, pendefinisian ROI, deteksi objek dengan menggunakan metode Faster R-CNN. Berdasarkan uji coba yang telah dilakukan pada penelitian ini, didapatkan nilai rata-rata presisi sebesar 98.21%, nilai rata-rata recall sebesar 97.09%, dan nilai rata-rata akurasi sebesar 95.36%.
================================================================================================
Coronavirus Disease 2019 (COVID-19) is a major health problem that causes acute respiratory disease in humans. The absence of immunity to COVID-19 increases susceptibility to being exposed to the virus, and the absence of approved drugs has resulted in efforts to control the transmission of the COVID-19 outbreak to rely heavily on non-pharmaceutical interventions such as individual prevention efforts, for example the application of health protocols. In 2020, WHO announced recommendations regarding the implementation of health protocols, one of which was the use of masks to prevent the transmission of the COVID-19 outbreak. However, this recommendation will be useless, if there is no awareness of each individual in the community to wear masks in accordance with existing procedures. From these problems, monitoring is needed to ensure the use of masks is applied in accordance with existing procedures. Monitoring can be done manually, but it is expensive and requires a lot of resources. Therefore, in this study an automatic monitoring system was designed by detecting mask users. In this study, there are three types of classes (labels) for the use of masks, namely Correct Use of Masks (Mask), Improperly Use of Masks (Improperly Masks), and No Masks (NoMask). The process of detecting the use of masks in this study was carried out at the testing stage, with the steps of video input, video acquisition, ROI definition, object detection using the Faster R-CNN method. Based on the trials that have been carried out in this study, the average value of precision is 98.21%, the average value of recall is 97.09%, and the average value of accuracy is 95.36%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Mask Usage, Faster R-CNN, Object Detection, Monitoring, Pemakaian Masker, Deteksi Objek, Monitoring.
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Muhammad Syahrul Romadhon
Date Deposited:	11 Feb 2022 03:42
Last Modified:	02 Nov 2022 01:25
URI:	http://repository.its.ac.id/id/eprint/93404

Actions (login required)

View Item