Volume 7, Issue 4, April – 2022 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Face Mask Detection System using

Open CV–Python and Machine Learning
Sonali, Prachi Sharma,
Guided by : Vidhu Jain

Abstract:- In this paper we are representing the The mask has a means to protect it from diseases like
methodology of face mask detection generation using corona. If you wear a mask, then the green colored rectangle
human. And through masks, people are protected from will sleep in my project.
epidemics like corona. And people are saved from this
epidemic In earlier times people used to use masks to And if you don't wear mask, then my project will read
keep their face beautiful, but in today's time people use your face and print no mask. In this research paper, we have
masks. To avoid disease like corona. And through this told how to solve the problem of face mask detection. And
paper, we ask to make people aware that everyone should how to check the mask. For this, we have made a program,
use masks and protect themselves from dangerous whose mask will be checked by using the camera. And those
diseases like corona. We have provided security to the who will not keep wearing masks and those who will
people through Face Mask Detection Project. And we continue to wear masks, they will be read on camera.
have used openasv-python in our project to read the faces
of the people. And we have used tensorflow package to To deal with this problem, we do not need any guard or
person who keeps an eye on the people. We can integrate a
read the mask in our project. And how many people have
put on masks and how many people have not, all this is camera that continuously clicks pictures of humans and from
there detects whether they are wearing a face mask or not.
read by my project.

Keywords:- Face mask detection system, machine learning, II. RELATED WORK
opencv-python, tensorflow package, deep learning,One-stage In face mask detection, we have used an algorithm so
detector, Two-stage detector. that we can easily read the mask on the face. And in face
I. INTRODUCTION mask detection, we have used some math matrix concepts.
With which we can easily read the face mask. And our
Face mask detection system refers to detect whether a system tells us which person mask is used and angle person
person is wearing a mask or not. If a person keeps on wearing mask is not used. If a person mask is not used, then my
a mask, then my project reads it and if a person has not system will print the mask and detect it in red color. And if a
applied a mask, then reads it and tells it. And through Face person is wearing a mask, then my system will detect the
Mask Detection System, we can find out which person has mask and print the face in green color.
not applied the mask and this is my face mask detection  The face mask detection system has been researched to read
system security based. The Face Mask Detection System the masks of the people. And a personal room is used in the
project is used to detect the face in a large company. And system from which the face mask of the people is easily
angle person has come with mask and angle person is not read. And the purpose of this research is to read the face
wearing mask: This can be easily known through my project. masks of the people. In the face mask detection system, 225
And to develop this project, we used machine learning and * 225 pixels read the face. System develops Raspberry Pi-
python and also used openasv-python. based real-time Facemask recognition that alarms and
captures faces.
Where the mask is detected using various machine  In this paper, measures have been taken to avoid diseases
learning algorithms for the purpose of security, like corona by taking face masks. And machine learning
authentication and surveillance. Face detection systems are a concept has been used in it. So that the face mask can be
major area in the field of computer vision. A significant body easily detected and disease like corona can be avoided. And
of research has contributed to algorithms for mask detection in this paper we want to tell about the activity of the people
in people. and how to avoid diseases like corona. If people come out
of the house, they must wear masks. Such people can avoid
And we have uploaded some image without mask and diseases like corona.
some image without mask in our project.  In this paper, we have included two proposed models. In
which the first is designed for convenience. And the second
We have used machine learning concept to read the
mask. And have used some important packages which have component has to be designed. Medical face masks are read
openasv-python, tensorflow, numpy, imutils, scipy, from Yolov2. And the two medical face mask data sets
matplotlib, keras. Face masks are easily read through these have been combined into one. And from this research the
packages. data set is examined.

The Numpy package is used to read the number of  Tensorflow:- In this paper we talked about TensorFlow
masks. FlowFramework. Tensorflow is an open source frame
block. Which is done in machine learning concept.And

IJISRT22APR1525 www.ijisrt.com 1156

Volume 7, Issue 4, April – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Tensorflow is a Python library. And the library is used to phone which includes deep learning. And during deep
develop it for big projects. And through this tensorflow learning, face recognition and speech recognition are done.
application can also be developed. And TensorFlow is Deep learning gives fast Deep learning thinks like humans.
called the core open source library. And this is done in That is why it is eaten like nooran. Hence it is often also
machine learning and deep learning. And we have done this called Deep Neural Learning and Deep Neural Networking.
frame work in the face mask detection system software.
Tensorflow is develop by google brain team. .
 Keras:- In this paper, we will also tell about the bra of
Keras library. keras is an open-source high-level neural
network library. This is done with the Python programming
language. It was developed by one of the Google engineers,
François Cholet. The TensorFlow library is used to control
the Keras library. Keras works on library network. It only
supports a combination of convolutional networks and
recurrent networks. Keras is an open source library. It has
been developed from the Python program. It is done with
machine learning and deep learning. When we develop a
project, then we link the Keras library in our project. With
which my project works on the network. And It is easy to
identify the user through the Keras library. Fig. 2: Deep Learning

 MobileNetV2 :- In this paper, we will tell about the use of

mobilenetv2. Today's time Mobilenetv2 provides more
power than the next time. It is a pleasure to tell us about it.
And Mobilenetv2 is an improvement of Mobilenetv1. this
mobile Object Detection, Semantic Segmentation and
advances mobile visual recognition. This is known as part
of mobilenetv2 tensorflow-slim image classification
library. Mobilenetv2 is used to read the face in the mobile.

Fig. 1: Keras logo

 OpenCV:- Opencv is a library for Python. Which is used to

detect the face. And this is done with machine learning.
And it is used to recognize his face. And using this library,
you can also detect the face of the animal. This is from the
concept of machine learning. And opencv represents an
algorithm. With which the face is easily recognized.
Openasv has more than 2550 computer vision and machine
learning algorithms, which are used in image and video
processing-related problems. It can be done with Python,
C++ and Java languages. And with it supports Windows,
Linux operating systems.
 Deep learning:- Deep learning is an artificial intelligence.
And it makes it read the person's brain with machine
learning. And the project of machine learning is developed
during deep learning. And deep learning is an important
learning. And deep learning is a subset of machine learning.
Deep leraning is known as Deep Neural.
 Learning or Deep Neural Network:- Deep learning uses
an indexed algorithm whereas machine learning uses a
linear algorithm. There is not much difference between
deep learning and machine learning, the work of both is the
same. By using deep learning, we can develop the biggest
projects. And in today's time, the biggest companies are
using Deep Learning. In which all the company of the

IJISRT22APR1525 www.ijisrt.com 1157

Volume 7, Issue 4, April – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Mobile Netv 2

Fig. 3: Mobilenet

 Dataset: We have used dataset created by Pragya Bhanwari in  with_out_mask :- The Without Masks folder has a total of
our project. And there are two folders in this dataset. 1800 images without masks. Images are detected through
 without_mask deep learning to people without masks.
 with_mask

Fig. 4: Without mask images

with_mask :- There is a total of 1700 images in the with mask folder. In which all the images are masked. The logo of the
mask is detected through deep learning. And through deep learning, it can be ascertained that the cone mask is applied and
the angle mask is not applied.

IJISRT22APR1525 www.ijisrt.com 1158

Volume 7, Issue 4, April – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig. 5: With mask images

 To find out the target, Pragya Bhanwari created the dataset and
the person who is wearing a mask detects it from the image of Next. We need the image to wear a mask. By which we
the mask. And those who have not applied a mask, match the find out the mask and match the mask or IMA with the one
image without a mask. This shows that the mask has been without the mask, so that we know who is wearing a mask or
applied. And the angle mask is not applied. who is not wearing a mask.
 Implementation:- We first read the face mask of the people,
If you use a set of images to create an artificial dataset of
then we detect the face of the people by using opencv .And
people wearing masks, you cannot "reuse" images without masks
then we use ROI to identify people by their eyes and mouth.
in your training set.
And from this we find out that the angle mask has been applied
and the mask has not been applied.

The structure of project.

Face-Mask-

├──

1. with_mask

2. with_out_mask

├── face_detector

1. deploy.prototxt

2. res10_300x300_ssd_iter_140000.caffemodel

├── detect_mask_video.py

├── mask_detector.model

├── plot.png

└── train_mask_detector.py
2 directories, 8

IJISRT22APR1525 www.ijisrt.com 1159

Volume 7, Issue 4, April – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
III. RESULTS

 When a person wears a mask, then the 'mask' comes through in my project. And it is seen in green color.

With_mask

Fig. 6: Code implementation mask detected

 When a person does not apply a mask, then in my project the 'ne mask' comes through white. And it is seen in red color

Fig. 7: Code implementation face mask not detected

IV. CONCLUSION REFERENCES

In this paper, we will tell about a project named Face [1.]https://www.simplilearn.com/keras-vs-tensorflow-vs-

Mask Detection System using Opencv-Python. And through pytorch-article#:
this project we can recognize people without masks. And this ~:text=TensorFlow%20is%20an%20open%2Dsourced,be
project can be done in a big company. And it can be used in cause%20it's%20built%2Din% 20Python.
the capacity room. And this project can also be done in the [2.]https://brain-mentors.com/face-mask-detection-using-
bank. And through this project we can digest how many opencv-in-python/
people have put on masks. And how many people have not [3.]I took help from my brother
put on masks. We have used openasv-python, tensorflowets [4.]https://data-flair.training/blogs/face-mask-detection-with-
package to create this project. And the dataset has also been python/
done to him. The total in the data set is 36009 images. In [5.]https://pyimagesearch.com/2020/05/04/covid-19-face-
which 1900 images are without masks. And 1700 image mask-detector-with-opencv-keras-tensorflow-and-deep-
masks have been applied. learning/
[6.]with the help of the teacher
[7.]we took the help of google
[8.]Took the help of the Geek for Geek website
[9.]I took help of youtube to write research paper

IJISRT22APR1525 www.ijisrt.com 1160