Facial Mask Detection System Based On COVID-19 Protocol

Volume 8, Issue 1, January – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Facial Mask Detection System based on

COVID-19 Protocol
Anisha Benjamin Chavan Dr. Vajid Khan
Dept. of Computer Engineering, Dept. of Computer Engineering,
K.J College of Engineering and Management Research, K.J College of Engineering and Management Research,
Pune, Maharashtra, India Pune, Maharashtra, India
Abstract:- A mask detection system which compares the II. METHODOLOGY

images of faces with and without masks, and alerts the
user to follow covid norms without any manual A. System Design
intervention using CNN-convolutional neural network Software design should be a method for converting
and facial recognition algorithm while creating an alert requirements into the appropriate representation on a regular
mechanism for the general public. In this report, in order basis. This graphic will be used to construct the structure and
to address the issue of the masked face region, we propose targeted plan diagrams, as well as the mission method.
a model based on discarding the masked region and deep
learning-based features. B. Existing System
Yolov3 is now being used to construct a system that is
Keywords:- Deep learning; CNN; MobilenetV, OpenCV. aware of face masks. They introduced a spine community that
will distribute more resources inside the current system, and
I. INTRODUCTION they used Glou and focus loss to speed up the coaching
process andimprove performance. It has an accuracy rate of
It has been almost 3 years that we are facing the wrath 86 percent. When using the appliance with one-of-a-kind
of COVID-19 all over the world affecting the lives of many algorithms, the accuracy will be doubled.
people. India is also not far away from this. India recorded
oneof the worst effects of COVID leading to the economical C. Proposed System
crisis. Lots of people lost their lives. Lockdown was imposed The mask detection in the proposed machine has been
leading to strain on the world economy. After the two waves done using MobileNet V2. In comparison to the previous
got over, relaxations were given. WHO gave instructions on system, MobileNet V2 will be able to detect face masks in a
how to take care of this virus. They said that masks are a great large group of men and women and provide greater accuracy.
weapon to keep coronavirus away. Proper sanitization should As a result of the entry dataset and therefore the segmented
be maintained whenever people go out. People after getting photo of the equation being received as output, provide an
relaxed in covid regulations started going out. Often we saw image of some persons wearing masks and now without
people not maintaining social distancing when going out in wearing masks. The mannequin is then performed employing
public places. As we know how important it is to maintain a camera, with the video being checked by way of body and
social distancing, sanitization, and wearing masks when we scaled as needed. The preprocessing feature is then used to
goout in public. After the revocation of covid rules, we have encourage the impacts of those wearing masks and those who
seenpeople going out in crowded places without masks and aren't.
maintaining social distance. Surveys tell that almost 90% of
people are aware whereas only 40 percent of people wear III. RESULTS & DISCUSSION
masks. From these surveys, we will say that people often
neglect to wear masks either because they are not comfortable The primary goal is to come up with an idea that can
or they don’t believe that there is such a thing called covid. motivate people to wear masks in public places by detecting
This kind of behavior by people leads to the quick spread of the people who are not wearing masks. In order to address
covid viruses rapidly affecting the lives of many people. The the issue of the masked face region, we are proposing a model
WHO has thus said that only wearing masks and maintaining based on discarding the masked region and deep learning-
social distancing is the key to keeping away the virus. The based features. We will detect the people who are not wearing
areas where people are freely moving without masks are the masks by eliminating the masked faces using the Neural
key culprits. Also, the people living in rural areas where there Network algorithms. After recognizing the facial features and
is no proper awareness, need to be aware of these facts and checking if the faces are with or without mask, an alert can
tips to keep away these viruses. This uncertain outbreak of be triggered to them for following the covid rules and
the COVID-19 virus has led us to the importance of wearing regulations.
masks in public places and keeping us safe from such viruses.
IJISRT23JAN172 www.ijisrt.com 865

ISSN No:-2456-2165
A) Image Segmentation using Mask CNN them, advising them to wear masks in public. We utilized
The Mask R-CNN is a CNN for image segmentation and Twilio for delivering SMS, which provides programmable
eventsegmentation that is built on top of a quicker R-CNN communication tools for sending and receiving text messages
that can come across objects and limits. Instance via its web service APIs, and the smtplib module for email,
segmentation or instance recognition is the process of which simply establishes an SMTP session for the client
detecting all objects in a photograph and segmenting each object and can then be used to send mail to another computer.
event. It emerges as a result of the harsh realities of object
detection, localization, and classification penalties. During
this method of segmentation, a clear separation between each
item classified as a similar circumstance is noted. Everyone
is treated as a single entity during the event segmentation
process. It's also known as foreground segmentation because
it works on the picture's subject matters rather than the
background-CNN and can produce two outputs for each
object, a class label, and a bounding field offset, whereas
Mask R-CNN can produce three outputs, including article
masks in addition to the class label and bounding container
offset. The more masks output isexclusive from the opposite
two outputs, implying that the finer the spatial design of an
item, the more masks output is required. Mask R-CNN is a
faster version of R-CNN that includes an output for object
masks in addition to current outputs such as classification
labels and bounding boxes.
Fig 2: General Frame of connection of the above modules
B) Implementing the Model in OpenCV
The model will be constructed with the help of a
IV. CONCLUSION
webcam, which reads the video frame by frame and resizes it
as needed. Then, in connection with the accuracy in %, the
In this mask detection, we used the MobilenetV2 set of
preprocessing feature is known as achieving the final result
rules to correctly distinguish persons wearing masks and
of human people wearing masks and now not wearing a mask. those without masks, as well as send an email to those
involved. Its overall performance in photographs is generally
C) Implementing the Face Recognition
accurate, and our detecting effects were also rather accurate.
The method of detecting a human face using technology
This detection will be utilized for video flow or digital
is known as facial reputation. Biometrics is used in a facial
digicam-fed inputs, as well. This can be used in places of work
reputation device to map face functions from a picture or
and establishments by way of training the database with
video. To find a match, it compares the information to a
personnel photos or students' photos and by way of face
database of known faces. reputation, and the character is diagnosed by way of a mobile
number that is unique for each person, and other information
Step 1: Detection of Face
about the individual is obtained from the database, and it will
Both by myself and in a crowd, the digital digicam
be simple to tell that specific character or beneficial for taking
recognizes and locates a snapshot of a face. The image might any actions concerned. The suggested version can be
also show the character searching ahead of time or in profile. improved by including other aspects such as the number of
persons and the social distance between individuals.
Step 2: Analysis of Face
Following that, a snapshot of the face is taken and
REFERENCES
examined. Because it may be simpler to integrate a 2D image
with publicphotographs or those in a database, most facial [1]. Xinbei Jiang, Tianshan Gao, Zichen Zhu, and Yukang
interest period is based on 2D as a potential rather than 3D Zhao., Real-Time Face Mask Detection Method Based
images. The gap between your eyes, the depth of your eye on YOLOv3, Electronics, 7, pp.130-147, 2021.
sockets, the space from the forehead to the chin, the form of [2]. Samuel Ady Sanjaya and Suryo Adi Rakhmawan., Face
your cheekbones, and the outline of your lips, ears, and chin Mask Detection Using MobileNetV2, International
are all crucial factors that the program interprets. The Journal of Engineering and Advanced Technology, 4,
objective is to learn the facial landmarks, which are perhaps pp.2249-8958, 2021.
the most important aspect of facial recognition. [3]. G. Jignesh Chowdary, Narinder Singh Punn, Sanjay
Kumar Sonbhadra, and Sonali Agarwal, Face Mask
D) Sending SMS and Email to Person who is notwearing Detection using Transfer Learning of InceptionV3,
Mask IEEE Access, 20, pp.456-665, 2021.
We're employing a face recognition system to identify [4]. Shilpa Sethi, Mamtha Kathuria and Trillok Kaushik.,
folks who aren't wearing masks, which will aid us in locating Face mask detection using deep learning, Multimedia
their information. We will send a preventive message in the Tools and Application, 8, pp.42-72, 2020.
form ofSMS and email to such persons in order to reward

ISSN No:-2456-2165
[5]. Chunli Qin, Demin Yao, Yonghong Shi and Zhijian
Song, “Computer-aided detection in chest radiography
based on artificial intelligence: a survey,” BioMedical
Engineering Online, vol. 17, pp. 1-23, 2018.
[6]. Riya Chiragkumar Shah and Rutva Jignesh Shah.,
Detection of Face Mask using Convolutional Neural
Network, Mobile Information System, 43, pp.382-487,
2019.
[7]. Safa Teboulbi, Seifeddine Messaoud, Mohamed Ali
Hajjaji, and Abdellatif Mtibaa., Real-Time
Implementation of AI-Based Face Mask Detection and
Social Distancing Measuring System for COVID-19
Prevention, Scientific Programming, 32, pp.167-254,
2021.
[8]. Xueping Su, Meng Gao, Jie Ren, Yunhong Li, Mian
Dong, and Xi Liu., Face mask detection and
classification through deep transfer learning,
Multimedia Tools and Applications, 53, pp.11042-
11772, 2021.
[9]. Mohamed Almghraby and Abdelrady Okasha Elnady.,
Face Mask Detection in Real-Time using MobileNetv2,
International Journal Of Engineering and Advanced
Technology, 6, pp.49-89, 2021
[10]. Chhaya Gupta and Nasib Singh Gill., Corona mask: A
Face Mask Detector for Real-Time Data, International
Journal of Advanced Trends in Computer Science and
Engineering, 9, pp.2278-3091, 2021.
[11]. Face Mask Dataset – Kaggle Repository.

Facial Mask Detection System Based On COVID-19 Protocol

Uploaded by

Copyright:

Available Formats

Facial Mask Detection System Based On COVID-19 Protocol

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Facial Mask Detection System Based On COVID-19 Protocol

Uploaded by

Copyright:

Available Formats

Volume 8, Issue 1, January – 2023 International Journal of Innovative Science and Research Technology

Facial Mask Detection System based on

Abstract:- A mask detection system which compares the II. METHODOLOGY

IJISRT23JAN172 www.ijisrt.com 865

IJISRT23JAN172 www.ijisrt.com 866

IJISRT23JAN172 www.ijisrt.com 867

You might also like