Skip to content
Home
All Articles
Recommended Courses
Real-Time Face Mask Detection
with Python
Aman Kharwal
December 27, 2020
Machine Learning
4
In the age of Covid-19, face masks and social distancing
have become the new normal. In this article, I will introduce
you to a machine learning project on real-time face mask
detection with Python.
Real-Time Face Mask Detection System
Indoor places, such as restaurants and grocery stores, are
legally required to have rules in place for the mandatory use
of face masks. Having a worker manually examining each
person to make sure their mask is on simply defeats the goal
of limiting contact with people as much as possible. So, a
real-time face mask detection system can be used to address
this issue that will not only maximize efficiency but will also
ensure to potentially save lives.
Also, Read – 100+ Machine Learning Projects Solved and
Explained.
The technology behind the real-time face mask detection
system is not new. In Machine Learning, face mask detection
is the problem of computer vision. Too often we see computer
vision applications of this technology in our daily lives. A
common example is a face unlocking in smartphones.
The goal of a face mask detection system is to create an
image recognition system that understands how image
classification works, and it should work with great accuracy
so that our model can be applied in the realtime situations. It
will work by recognizing the boundaries of the face and
predicting whether or not you are wearing a face mask in
real-time.
Real-time Face Mask Detection with Python
Now, I’m going to create a convolutional neural network to
create a real-time facial mask detection model with Python.
Here, I will use three dense layers in our model with
respectively 50, 35 and finally 2 neurons. The dense network
produces the probability of the binary classification of no
mask = 1 and mask = 0:
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dropout, Dense
cnn = Sequential([Conv2D(filters=100, kernel_size=(3,3),
activation='relu'),
MaxPooling2D(pool_size=(2,2)),
Conv2D(filters=100, kernel_size=(3,3),
activation='relu'),
MaxPooling2D(pool_size=(2,2)),
Flatten(),
Dropout(0.5),
Dense(50),
Dense(35),
Dense(2)])
cnn.compile(optimizer='adam', loss='binary_crossentropy', metrics=['acc'])
view rawreal-time facemask detection.py hosted with ❤ by GitHub
Testing The Model in Real-Time
To test our model in real-time, I’ll be using the VideoCapture
function in the OpenCV library in Python. The Cascade
classifier, designed by OpenCV, was used to detect the
frontal face in live video via detectMultiScale. We can use a
while loop to continue capturing images from the webcam.
Our machine learning model will then determine whether or
not a face mask is worn in real-time. Based on the
performance and accuracy of our model, the result of the
binary classifier will be indicated by showing a green
rectangle superimposed around the section of the face
indicating that the person at the camera is wearing a mask, or
a red rectangle indicating that the person on camera is not
wearing a mask.
Now let’s implement the above steps for testing the face
mask detection model in real-time by using the Python
programming language:
import cv2
import numpy as np
labels_dict={0:'No mask',1:'Mask'}
color_dict={0:(0,0,255),1:(0,255,0)}
imgsize = 4 #set image resize
camera = cv2.VideoCapture(0) # Turn on camera
# Identify frontal face
classifier = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
while True:
(rval, im) = camera.read()
im=cv2.flip(im,1,1) #mirrow the image
imgs = cv2.resize(im, (im.shape[1] // imgsize, im.shape[0] //
imgsize))
face_rec = classifier.detectMultiScale(imgs)
for i in face_rec: # Overlay rectangle on face
(x, y, l, w) = [v * imgsize for v in i]
face_img = im[y:y+w, x:x+l]
resized=cv2.resize(face_img,(150,150))
normalized=resized/255.0
reshaped=np.reshape(normalized,(1,150,150,3))
reshaped = np.vstack([reshaped])
result=cnn.predict(reshaped)
label=np.argmax(result,axis=1)[0]
cv2.rectangle(im,(x,y),(x+l,y+w),color_dict[label],2)
cv2.rectangle(im,(x,y-40),(x+l,y),color_dict[label],-1)
cv2.putText(im, labels_dict[label], (x, y-
10),cv2.FONT_HERSHEY_SIMPLEX,0.8,(255,255,255),2)
cv2.imshow('LIVE',im)
key = cv2.waitKey(10)
# stop loop by ESC
if key == 27: # The Esc key
break
webcam.release()
cv2.destroyAllWindows()
view rawreal-time facemask detection.py hosted with ❤ by GitHub
https://www.youtube.com/watch?v=whNqjwvFhiI
So you can see that our model gave good results and it can
be used in real-time scenarios. Hope you liked this article on
machine learning project on Real-Time Face Mask Detection
with Python. Please feel free to ask your valuable questions
in the comments section below.
Aman Kharwal
Data Strategist at Statso. My aim is to decode data
science for the real world in the most simple words.
ARTICLES: 1714
