Master of technology in information technology department of information science and

---

i 
REAL-TIME FACE MASK DETECTION USING
MOBILENET V2 
PHASE 1 REPORT
 
Submitted by
JANANI A 
(2020246030) 
in partial fulfillment for the award of the degree of
MASTER OF TECHNOLOGY IN 
INFORMATION TECHNOLOGY
 
 
DEPARTMENT OF INFORMATION SCIENCE AND 
TECHNOLOGY 
COLLEGE OF ENGINEERING, GUINDY
ANNA UNIVERSITY
CHENNAI 600 025
FEB 2022

---

ii 
ANNA UNIVERSITY
, 
CHENNAI 
BONAFIDE CERTIFICATE 
Certified that this Report titled 
"REAL-TIME FACE MASK DETECTION 
USING MOBILENET V2”
for the Dissertation – I of the project work is the bonafide 
work 
JANANI A (2020246030)
who carried out the work under my supervision. 
Certified further that to the best of my knowledge the work reported herein does not form 
part of any other thesis or dissertation based on which a degree or award was conferred 
on an earlier occasion on this or any other candidate. 
 
PLACE: 
Chennai
Dr. M.VIJIYALAKSHMI
DATE :
 
 
 
 
 
 
ASSOCIATE PROFESSOR
PROJECT GUIDE
DEPARTMENT OF IST, CEG 
ANNA UNIVERSTY 
CHENNAI 600025 
COUNTERSIGNED 
 
Dr. S. SRIDHAR
HEAD OF THE DEPARTMENT
DEPARTMENT OF INFORMATION SCIENCE AND TECHNOLOGY 
COLLEGE OF ENGINEERING, GUINDY
ANNA UNIVERSITY
, 
CHENNAI 600025 

---

iii 
ABSTRACT
Today the biggest problem the world is facing above all the natural disasters 
is the Covid-19. It's been more than a year but the solution to the issue is still at a 
far fetch. However, still have few ways to control the outbreak as instructed by the 
WHO (World Health Organization). A few among them are wearing a mask and 
maintaining social distance. The objective of the project is to detect face masks in 
a public gathering or an event. The algorithm used in the project to achieve the 
objective is MobileNet V2. An image of a few people wearing a mask and without 
wearing a mask is used as an input dataset. There are few processes involved in 
achieving the objective of the project that include pre-processing, data 
augmentation, training, testing and image segmentation. After the processes, with 
the help of the Mask R-CNN algorithm, will get a segmented image of the input 
dataset of people wearing a mask and people not wearing a mask. Then, the model 
is implemented using a webcam where get result of people wearing a mask and not 
wearing a mask along with the accuracy in percentage. In this project, developing a 
model that can detect people who are not wearing masks in public places. This 
project can be merged with real time applications at airports, railway stations, 
workplaces, schools, and other public places to ensure compliance with the 
guidelines for public safety. In the middle of the Covid-19 crisis the project 
focuses on achieving the guidelines provided by WHO to control the spread of 
Covid-19.

---

iv