Journal of Applied Informatics and Computing (JAIC)
Vol.6, No.1, Juli 2022, pp. 01~06
e-ISSN: 2548-6861 1
http://jurnal.polibatam.ac.id/index.php/JAIC
Implementation of YOLO-v5 for a real-time Social Distancing Detection
Falah Hikamudin Arby
1
*, Imam Husni Al Amin
2
*
* Teknik Informatika, Universitas Stikubank Semarang
falahhikamudin.arby@gmail.com
1
,imam@edu.unisbank.ac.id
2
Article Info
ABSTRACT
Article history:
Received 2021-10-13
Revised 2021-12-17
Accepted 2022-02-14
The world is in an uproar with the Covid-19 pandemic, which has impacted
society. Governments have made various efforts worldwide to suppress the spread
of the Covid-19 virus. One of the health protocols that the government has
appealed is social distancing or social restrictions, namely limiting interactions
between human beings as long as 1-2 meters. However, in reality, many people still
ignore social distancing policies. The application of a social distancing detection
system can solve this problem. This system aims to detect people who are violating
health protocols in the form of social distancing and then issue a voice warning to
keep their distance from others to avoid the spread of the Covid-19 virus by using
the YOLO-v5 method, which is the latest version of YOLO (You Only Look
Once). The processing speed of YOLO-v5 has increased drastically, with the
fastest speed reaching 140 Frames Per Second (FPS). It has a small size, even
having a size of 90% compared to the previous version. The accuracy of human
detection using YOLO-v5 from this system reaches 83.28%, and the accuracy of
social distancing detection reaches 90.8%. From the results of the percentage
analysis carried out, it can be concluded that the system created can function well
for social distancing detection. However, it is not easy to detect if humans are too
far from the camera.
Keyword:
Computer Vision,
Covid-19,
Deep Learning,
Social Distancing,
YOLO.
This is an open access article under the CC–BY-SA license.
I. INTRODUCTION
The Coronavirus disease 2019 (Covid-19) outbreak is a
phenomenon of a deadly disease outbreak that has spread
throughout the world, including in Indonesia. Health
workers around the world are currently making efforts to
control the outbreak of this disease which was emerged first
in Wuhan, China, at the end of 2019 [1]. To date, the World
Health Organization (WHO) on August 31, 2021, reported
that this virus infected 215,714,824 people, with 4,490,753
deaths worldwide [2].
The Covid-19 virus has a high level of infection to be
transmitted to people around it. It is known that the main
route of the spread of this virus is through liquid droplets
that attack the respiratory tract [3]. Therefore, both in
Indonesia and worldwide, the government urges the public
to implement health protocols, including wearing masks
when travelling, constantly washing hands, and
implementing physical distancing and social distancing.
Social distancing is one of the steps that the government
has taken to reduce the risk of spreading the Covid-19 virus.
The intended social restrictions are not shaking hands and
limiting a distance of at least 1-2 meters when interacting
with other people [4]. The goal is to avoid liquid droplets
from someone at high risk of suffering from Covid-19 to
avoid getting infected. However, in reality, not a few people
do not obey the rules of this social distancing health
protocol. Many people ignore their distance when interacting
with each other. This has become a concern for both the
government and other communities because, of course, this
has the potential to spread the Covid-19 virus.
Deep Learning has become a hot topic in recent years,
some of which are used to create objects, faces, and other
types of detection. Several popular detection methods used,
such as Fast-RCNN, Faster-RCNN, SSD (Single Shot
Multibox Detector), and You Only Look Once (YOLO),
have become an evolution of precise detection methods but
light in use in several aspects [5].
资源评论
