Social Distancing Monitoring With Alarm System For Limited Face To Face Classes in Covid 19

St.
Paul University Surigao

St. Paul University System
Surigao City, Philippines
SOCIAL DISTANCE MONITORING WITH ALARM SYSTEM FOR LIMITED

FACE TO FACE CLASSES IN COVID-19
A Thesis Proposal Presented to

The College of Engineering
St. Paul University Surigao
In Partial Fulfillment
of the Requirements for the Degree
BACHELOR OF SCIENCE IN COMPUTER ENGINEERING
By:
FRANCIS RAE R. CILLO

MARK JASON M. CONTE
DRAKE JOSHUA J. BESARIO
May 2022
`
APPROVAL SHEET
This thesis entitled
SOCIAL DISTANCE MONITORING WITH ALARM SYSTEM FOR LIMITED

FACE TO FACE CLASSES IN COVID-19
prepared and submitted by Francis Rae R. Cillo, Mark Jason M. Conte and
Drake Joshua J. Besario has been examined and is recommended for approval and
acceptance for
ORAL EXAMINATION.
Engr. Angelus Vincent P. Guilalas

Adviser
----------------------------------------------------------------------------
PANEL OF EXAMINERS
APPROVED by the Committee of Oral Examination with a grade of _____
on __________________.
Sr. Marie Rosanne B. Mallillin, SPC

Chair
_______________________________ _______________________
Member Member
ACCEPTED in partial fulfillment of the requirements for the degree of

BACHELOR OF SCIENCE IN COMPUTER ENGINEERING
___Engr. Angelus Vincent P. Guilalas___

Dean, College of Engineering
`
Certificate of Originality
This is to certify that this research paper is our work and does not contain any
material previously published or written by another person nor material which to a
substantial extent has been accepted for the award of any degree of diploma in Saint Paul
University Surigao, and other educational institutions, except where due acknowledgment
is made in the thesis. Any contribution made to the research by others, with whom we
have worked at Saint Paul University Surigao or elsewhere, is explicitly acknowledged in
the thesis.

Candidate
MARK JASON M. CONTE

Candidate

Candidate
Date: _________________
Attested by:
Engr. Angelus Vincent P. Guilalas
Adviser
Date: __________
`
ACKNOWLEDGEMENT
The researchers would like to express the following individuals for their
assistance to the success of this thesis. First and primarily, the God Almighty, for his
endless support and encouragement, offering them with strength and perseverance
through hard moments, as well as wisdom and understanding to ensure the success of this
project and providing them with the necessities; Their parents for their unending support,
particularly in terms of cash, love and understanding, counsel and messages of support
that bolstered their spirits to continue pushing despite the struggles and problems they
endured;
Engr. Angelus Vincent Guilalas, research adviser and research instructor, for his
noteworthy and effective teaching in developing this project and for his teaching about
the process and for motivating the researchers for success which greatly helps throughout
the research. To their beloved friends, who extended their support by sharing their
knowledge, ideas and suggestion; To the faculty of Saint Paul University Surigao and to
those who were not mentioned but have directly and indirectly contributed their
assistance; To all of you, the researchers are thankful.
MARK JASON M. CONTE
`
ABSTRACT
Social distancing throughout to be and adequate precaution against the spread of
the pandemic virus. The risks of virus spread can be minimized by avoiding physical
contact among people.
The purpose of this project is, to produce a system for social distance monitoring
using an overhead perspective combining the advanced detection algorithms, this method
can be implemented in camera by the use of Pycharm, an Integrated Development
Environment (IDE) and Python 3.10 language. During this way, the detection uses pre-
trained algorithm using an overhead human data set. We used an approximate distance of
50 (N pixel) and set a threshold. A violation threshold is established to identify whether
or not the distance value breaches the minimum social distance threshold, additionally, a
tracking algorithm is employed to detect individuals in video sequences, specified the one
that violates and crosses the social distance threshold is also being mark and send an
alarm sound through the speaker. Experiments are distributed video sequences to check
the accuracy, precision and sensitivity of the system. Findings indicate that the developed
system successfully distinguishes the accuracy of the program in detecting the social
distance violation is 78.5%
Keywords: Computer vision, Person detection, Algorithm, Pycharm (IDE), Python
3.10
`
TABLE OF CONTENTS
Page
TITLE i
APPROVAL SHEET ii
CERTIFICATE OF ORIGINALITY iii
ACKNOWLEGDGMENT iv
ABTRACT v
TABLE OF CONTENTS vi
CHAPTER 1
THE PROBLEM AND ITS BACKGROUND 9

Conceptual Framework of the Study 11
Statement of Purpose and Objectives 14

Assumptions 14
Significance of the Study 15
Scope and Limitation 16
Definition of Terms 17
CHAPTER 2
REVIEW OF RELATED LITERATURE 20
Synthesis of the Review 23
CHAPTER 3
METHOD 24
Research Design 24
Hardware Design 25
`
Software Design 26
Data Gathering Procedure 31
Setting 33
Statistical tool 34
CHAPTER 4
RESULTS AND DISCUSSION 38
Technical Specification 38
Result 41
CHAPTER 5
Summary 54
Findings 55
Recommendation 55
Conclusion 57
REFERENCES 58
APPENDICES
A Letter for Adviser 61
B Letter of the Request to Conduct the Study 62
C Program code 63
CURRICULUM VITAE 76
LIST OF FIGURES
FIGURES
1. Conceptual Diagram 12
2. Flowchart of the study 13
3. System Overview 25
4. Person Detection 26
5. Distance Measuring 27
6. Euclidean Diagram 28
`
7. Distance Calculation 28
8. Flowchart of the system 30
9. Data gathering 32
10. Scenario 1> meter 41
11. Scenario <1 meter 44
12. Classroom conducted project 49
13. Monitoring Social distance in the conducted area 50
LIST OF TABLES
1. Table 1 Hardware components 38
2. Table 2 System Input 39
3. Table 3 Scenario of >1 meter 43
4. Table 4 Scenario of <1 meter 46
5. Table 5 Accuracy Calculation 47
6. Table 6 Precision Calculation 48
7. Table 7 Recall Calculation 48
8. Table 8 Accuracy Calculation 51
9. Table 9 Precision Calculation 52
10. Table 10 Recall Calculation 53
`
CHAPTER 1
THE PROBLEM AND ITS BACKGROUND
COVID-19 originated from Wuhan; China has affected many countries
worldwide since December 2019. On March 11, 2020, the World Health Organization
(WHO) announced a pandemic disease as the virus spread through 114 countries, caused
4000 deaths and 118,000 active cases (WHO; W.H. Organization, 2020). On October 7,
2020, they reported more than 35,537,491 confirmed COVID-19 cases, including
1,042,798 deaths. According to data obtained by the World Health Organization (WHO),
the world pandemic of COVID-19 has severely impacted the planet and has now infected
over eight million people worldwide. Wearing face masks and following safe social
distancing are two of the improved safety protocols that require to be followed publicly
`
10
places so as to stop the spread of the virus. to make a secure environment that contributes
to public safety (World Health Organization, 2020). [1]
In the Philippines, according to the Department of Health (DOH), the statistics of
the number of cases and death in COVID-19 are increasing due to the people who did
not follow the safety protocols, following the announcement of limited face to face class
at all education levels.
Commission on Higher Education (CHED) proposed the limited face-to-face class to
other degree programs that require hands-on experience in Higher Education Institutions
(HEIs)
The urgent need for hands-on experience and the safety of students taking up medicine
and allied health courses, who were permitted by the national government to conduct
limited face-to-face classes since January 2021, is the key for CHED to recommend on
the expansion proposal. [2]
The designed project Social Distance Monitoring with Alarm for the limited face-
to-face class is a system to distinguish between individuals to make sure social distance is
being followed and to slow down the spread of infectious diseases. In this project, the
researchers will implement social distancing monitoring, the researcher will apply object
`
11
detection (for person class) to detect all people and then apply distancing measures (N
pixels.) to monitor the distance between people. Also, by putting an alarm for the system
to notify the people for violating the minimum distance.
This designed project is to help the institutions to minimized the crowd in a
specific area or room and prevent getting contact with a virus and to make use of this
project for the future.
Conceptual Framework of the Study
The proposed project helps to ensure the student's safety by automatically monitoring
whether they maintain social distance. It anchors on the use of computer vision for
human detection in a video sequence and uses a triangular similarity technique to
measure the distance between persons detected by the camera.
The overall conceptual diagram is presented in work in Figure 1 is divided into
two modules, person detection and social distance monitoring. The top view installed a
camera capturing images from space and processed to the human detection module; the
human detection module utilizes detects human bounding boxes. The bounding box
`
12
information is further utilized by a social distance monitoring module that identifies the
violations and further processed it to the system unit. [3]
SYSTEM
Camera Input Frame
Detect violations and mark red PersonDetection
`
13
Speaker alarm output
Figure 1. Conceptual Diagram
Figure 1. Ddisplays the flow of the study. The input box represents the list of
materials and tools being used including the needed data. To identify individuals, the
design project employs a camera and a personal computer loaded with Pycharm, an
Integrated Development Environment, and Python 3.10 as a programming language, as
well as a speaker that emits a beep sound whenever people exceed the safe distance.
INPUT PROCESS OUTPUT
Hardware
Planning and
Requirements:
brainstorming
Laptop or Personal
Social Distance
Computer Selection of algorithm
Monitoring With
Web Camera
Determining the Scope
Speaker Alarm System For
of the Study
Software
Limited Face to
Requirements: Recognition of the
Face Classes in
Pycharm (IDE) Problems and
Covid-19
Language: limitations
Python 3.10 Development of System
Data:
` Camera
14
Performance of the Design

in terms of:
Accuracy Precision
Sensitivity
Figure 2. Flow of the study
Statement of Purpose and Objective
This project is to help the institutions to observing social distancing. As a result of
this, we are developing a system that creates the use of Pycharm as an Intergrated
Development Environment (IDE) and python 3.10 programming language to achieve the
system in social distancing monitoring with alarm.
Specific Objective:
1. To design and develop a system that monitors the social distancing during the
conduct of limited face to face class
`
15
2. To design and develop a social distancing system using Pycharm and Integrated
Development Environment (IDE) based on python program language.
3. To evaluate the performance of the design project based on different scenarios in
terms of precision, accuracy and sensitivity using confusion matrix.
4. To evaluate the system’s performance and test the design project to a school
where actual face-to-face class implemented.
Assumptions
For the Social Distancing Monitoring System to efficiently work there are crucial
assumption were made.
In the process of implementing the designed project the researcher assumed that:
• Develop a real-time based detection system.
• Accurate result of social distancing detection.
• Accessible components and resources.
• Achieve the monitoring social distance system with alarm.
Significance of the Study
This research will provide deep learning method and an advanced way for
monitoring for the safety of any individuals and to ensure the safety protocols in this time
of the pandemic, and this research will be beneficial to the following:
`
16
Educational Institutions: Due to the global immunization, some of the area with a
decreasing number of virus cases is announcing to open the schools for limited face to
face classes. In helping to ensure the safety of the students this research will help
monitoring the students to always follow the safety protocols, and to prevent possible
outbreaks inside the schools.
Public Establishment: Due to the lightening of the quarantine many establishments are
now open for business, this will also beneficial to monitor the crowd for a specific
establishment to always follow precautionary measures and because of its automated
system this will inform not only individually but the people in the entire area.
Government Agencies: This research will also help in the contact tracing; agencies will
easily know where, when and how the outbreak started and also due to its surveillance
system it can be install in any specified location to monitor the pedestrians and any
overcrowded places.
Community: One of the major priorities of this research is to help the people to always
ensure their safety in any kinds of activity, and to remind that even the quarantine is
minimized the virus is always there, people need to always follow what are the safety
protocols implemented by the government.
Scope and Limitation of the Study
`
17
This study is focused on a system that will automatically alert when someone is
not observing social distancing. The major operation and development system is to help
people and to minimize the number of infected viruses.
The Social Distance Monitoring with Alarm System in Limited Face to Face
Classes for COVID-19 is composed of different Python packages and Pycharm as an
Intergrated Development Environment which mainly concentrates on the operation in
programming the system.
This project is restricted to the campus area, using the camera it will monitor an
individuals for not observing the social distancing.
The designed project will only detect the persons violating social distancing, it
will not recognize any object or an animal, it only applies to humans. This project will be
utilized in future work likewise.
Definition of Terms
For a greater understanding of this study, the following terms are defined in the
context of this research.
`
18
Monitor - A computer monitor is an output device that displays information in pictorial or
text form. A monitor usually comprises a visual display, some circuitry, a casing, and a
power supply. [4]
Social Distancing - in public health, social distancing, also called physical distancing, is
a set of non-pharmaceutical interventions or measures intended to prevent the spread of a
contagious disease by maintaining a physical distance between people and reducing the
number of times people come into close contact with each other. [5]
Algorithms - In mathematics and computer science, an algorithm is a finite sequence of
rigorous well-defined instructions, typically used to solve a class of specific problems or
to perform a computation. Algorithms are used as specifications for performing
calculations and data processing. [6]
System - A system is a group of interacting or interrelated elements that act according to
a set of rules to form a unified whole. A system, surrounded and influenced by its
environment, is described by its boundaries, structure and purpose and expressed in its
functioning. [7]
Framework - an essential supporting structure of a building, vehicle, or object. [8]

Pixel - a minute area of illumination on a display screen, one of many from which an
image is composed. [9]
`
19
Computer vision - Computer vision is an interdisciplinary scientific field that deals with
how computers can gain high-level understanding from digital images or videos. From
the perspective of engineering, [10]
Real-time - the actual time during which a process or event occurs. [11]
Computer - an electronic device for storing and processing data, typically in binary
form, according to instructions given to it in a variable program. [12]
Alarm system - Alarm system means a device or series of devices, which emit or
transmit an audible or remote visual or electronic alarm signal, which is intended to
summon Police response. [13]
COVID-19 - Coronavirus disease (COVID-19) is an infectious disease caused by the
SARS-CoV-2 virus. [14]
`
20
CHAPTER 2
REVIEW OF RELATED LITERATURE
Social Distancing or also called as Physical Distancing, primarily means keeping
a safe distance between yourself and other people who aren’t from your household. The
virus cases increase among people who are in close contact for a long period of time, the
spread happens when an infected person has cough, cold and other flu-like symptoms as
the droplets from their mouth or nose launched into the air and lands into the face of
people nearby and for that reason why masks is important but masks alone will never
protect against the virus. To maintain or practice social distancing, people should stay at
least 1 meter away from other people and avoid contacting people in crowded places.
[15]
`
21
Person detection
Based on the study of the problem of detecting and tracking people in crowded
video scenes. While detection of individual objects has greatly improved in recent years,
crowd scenes remain particularly difficult for detection and tracking tasks due to heavy
occlusions, high person densities, and significant variation in people's appearance. To
address these issues, proposed leveraging information about the overall structure of the
scene and resolving all detections jointly. Investigated crowd density constraints in
particular, and formulate person detection as the optimization of a joint energy function
combining crowd density estimation and individual person localization. We demonstrate
how the optimization of such an energy function significantly improves person detection
and tracking in crowds. [16]
Surveillance and Monitoring
According to under the three-year Video Surveillance and Monitoring (VSAM) project
(1997--1999), the Robotics Institute at Carnegie Mellon University (CMU) and the
Sarnoff Corporation developed a system for autonomous Video Surveillance and
Monitoring. The technical approach uses multiple, cooperative video sensors to provide
continuous coverage of people and vehicles in a cluttered environment. This final report
`
22
presents an overview of the system, and of the technical accomplishments that have been
achieved. [17]
Alarm System
In the first stage, an approximate median method is used to segment moving regions in a
video frame. In the second stage, a fuzzy c-means (FCM) method is used to cluster
candidate smoke regions from these moving regions. In the third phase, a parameter
extraction method is used to extract a set of parameters from spatial and temporal
characteristics of the candidate smoke regions; these parameters include the motion
vector, surface roughness and area randomness of smoke. In the fourth stage, the
parameters extracted from the third stage are used as input feature vectors to train
a support vector machine (SVM) classifier, which is then used by the smoke alarm to
make a decision. Experimental results show that the proposed four-stage smoke-detection
algorithm outperforms conventional smoke-detection algorithms in terms of accuracy of
smoke detection, providing a low false-alarm rate and high reliability in open and large
spaces. [18]
Social Distance
An artificial intelligence network-based smart camera system prototype has been built in
this study, which measures social distance from a bird's-eye perspective. To recognize
`
23
persons in video sequences, the "MobileNet SSD-v3", "Faster-R-CNN Inception-v2", and
"Faster-R-CNN ResNet-50" models have been used. The final prototype, which is based
on the Faster R-CNN model, is an integrated embedded system that uses the camera to
measure social distance. The program, which was created with the "Nvidia Jetson Nano"
development kit and the Raspberry Pi camera module, does all necessary calculations,
identifies social distance violations, issues auditory and visual warnings, and sends the
findings to the server. The designed smart camera prototype is expected to be integrated
into public places as part of the world's transition to "sustainable smart cities." [19]
Synthesis of the Review
The above literature discussed about social distancing for COVID-19 that provides
information to the proponents that their proposed study that has similarities with other
system which is widely used. The ideas presented by the cited authors give the
researchers pattern for the architecture and algorithm of the study. As stated, the system
discussed in the readings is like that of this study in terms of the functions of the system.
They also shared the same idea in the purpose for which the system was made, that to,
automatically monitor social distancing with alarm. They differ in application used and
implementation of the program.
`
24
CHAPTER 3
METHODOLOGY
To identify Social Distance Monitoring with Alarm System in Limited Face to
Face Classes for COVID-19, this study will utilize a descriptive research approach.
Descriptive study attempts to correctly and methodically characterize a population,
situation, or phenomena, and the development of a designed system for all the people of
our society who wish to engage in social distance.
RESEARCH DESIGN
This study used of descriptive and developmental research design since the design
considered necessary for testing, observing without changing the quantitative data and
`
25
documenting the results. Also, the researchers will employ surveys in the area where a
school started the limited Face-to-Face classes gather needed information.
Project Developmental method
The study is a developing system for social distance monitoring by using Python
programming language with different python package and libraries, Pycharm IDE to the
camera to be able to detect people individual distance and with an alert sound, to politely
remind people to maintain their social distance.
Hardware Design
The hardware design is the physical and external composition of the system. It consisted
of the components that were interconnected and used in the project design.
SYSTEM OVERVIEW:
Figure 3. System Overview
`
26
Assuming that two people are in close proximity. If they cross the acceptable
social distance threshold our camera will respond by sending the detected people to the
System, where our code resides. The code will wait for the moment when the social
distance threshold has been violated and immediately sound an alert to a loud speaker,
either via the analog cable or HDMI to an external speaker. [20]
Software Design
The proponents used a Python, Pycharm and camera to capture people in specified area,
the image is parsed to detect people in the image. Upon detection the depth info is also
added to the detection.
`
27
Person Detection
Fig. 4. Person Detection
Looking in Figure 1 We can see here that if the camera will capture any individual person
it will notify the system and it will automatically marked as a person by putting a
rectangular shape. This is important for the system to know that it will automatically
recognize a person.
Distance Measuring
In Figure 5 It’s easier to imagine the position of people in stick figure, taking only the X
and Y coordinates into account. This perspective makes it possible to use a Euclidean
`
28
formula for distance calculations and can be useful to understand how the camera is
determining out position in space. Knowing each person’s position, we can now calculate
the distances between them and check if they do not exceed the defined threshold.
1meter
Distance
x,y x,y
Distance Calculations
Euclidean Formula
Fig.5 Distance Measuring
Frame 1
`
29
Fig 6. Euclidean diagram
In Fig 6 Frame 1, Assume this is a video frame captured by a camera; the algorithm will
create a rectangle box, with the 2 centroid measuring the distance between a human and
the camera.
Figure 7. Distance Calculation
`
30
In the next frame 2 the purple circle represents the previous centroid, while the yellow
circle represents the new centroid. If the yellow person moves closer to the purple person,
we can use the Euclidean formula to calculate the distance between them.
To compute the distance between the 2 centroid of the bounding box, using Pycharm and
a pre-built package of Euclidean formula to measure the distance between two bounding
boxes. The social distancing measurement in the video frame detected by the camera is
50 pixels.
`
31
Figure 8. Flow Chart of the system
Figure 2 shows the flow of the project design, it first access video stream from the
camera, ensures that at least two people detected in the video in order to compute
`
32
pairwise distance. If no one is violating social distance, a green bounding box will
appear; if someone is violating social distancing a red bounding box will appear along
with a beep sound from the speaker.
DATA GATHERING PROCEDURE
This section shows the flow of gathering data relevant on the research study. The
researchers presented a formal letter of request asking consent to conduct a study in a
school where face to face classes is implemented. Upon approval of the said request, the
researchers conducted an observation within the room by setting-up the computer and the
camera and with the use of the system the researcher can now gathered and analyze the
data.
`
33
STARTT
REQUEST FOR APPROVAL

TO CONDUCT STUDY
REQUEST INSTALLATION
OF CAMERA
IMPLEMENTATION OF THE
PROJECT
TESTING PROJECT
PERFORMANCE
VALIDATION OF RESULTS
END
Figure 9. Data Gathering Flowchart
`
34
SETTINGS
In choosing the setting of this study, the researchers will choose a school where
the limited-face-to-face-classes are implemented.
DATA ANALYSIS
The following were the data’s needs to be processed and also used to help in the analysis:
Project Development Tool
Block Diagram. It consisted of the components that were interconnected and used in the
project design.
Flowchart. This shows the flow of the designed project on processing the data from the
sensors used.
`
35
Statistical Tool
Confusion Matrix
A confusion matrix is a table that is used to define the performance of a classification
algorithm. A confusion matrix visualizes and summarizes the performance of a
classification algorithm.
Predicted
Actual Negative Positive
Negative TN FP
Positive FN TP
Confusion matrices represent counts from predicted and actual values. The output “TN”
stands for True Negative which shows the number of negative examples classified
accurately. Similarly, “TP” stands for True Positive which indicates the number of
positive examples classified accurately. The term “FP” shows False Positive value, i.e.,
the number of actual negative examples classified as positive; and “FN” means a False
Negative value which is the number of actual positive examples classified as negative.
One of the most commonly used metrics while performing classification is accuracy. The
`
36
accuracy of a model (through a confusion matrix) is calculated using the given formula
below.
Accuracy = TN+TP
TN+FP+FN+TP
There are concerns however with accuracy. It assumes equal costs for both
kinds of errors. Depending on the issue, 99 percent accuracy can be excellent,
decent, average, bad or horrible.
Recall:
A recall can be defined as the ratio between the total number of positive
examples correctly classified and the total number of positive examples. High Recall
suggests the proper recognition of the class (small number of FN).
Recall is given by the relation:
`
37
Precision:
To have the value of accuracy, we divided the total number of correctly
classified positive examples by the total number of estimated positive examples.
High accuracy demonstrates an example classified as positive, indeed positive (small
number of FP).
Precision is given by the relation:
High recall, low precision: This means that most of the positive examples are
correctly recognized (low FN) but there are a lot of false positives.
Low recall, high precision: This illustrates that a lot of positive examples (high FN)
are missing, but those we expect are actually positive ones (low FP)
F-measure:
Because we have two measurements (Precision and Recall), having one that
reflects both of them is helpful. We compute an F-measure that uses Harmonic Mean
instead of Arithmetic Mean because it punishes extreme values more. [21]
`
38
The F-Measure Has always been closer to the lesser value of Precision or Recall.
CHAPTER 4
`
39
RESULTS AND DISCUSSION
This chapter presents the documented design project based on input, process and output.
Technical Specification
Hardware Components
TABLE 1
Components Specification Picture
Computer Processor: AMD

Ryzen 3 3200G
System
RAM: 4GB DDR4
Camera A4 Tech Webcam
`
40
Speaker Genius wired speaker
Monitor Acer 60Hz
`
41
Coding Python 3.10 Libraries Output
Pycharm Integrated Numpy Social Distance Monitoring

Development Environment with Alarm System
(IDE) Cv2
Argparse
Imutils
Os
Scipy
Distance
Yolo
TABLE 2
System Input
The design project is powered by a computer with installed camera and speaker. The
system is controlled by Pycharm (IDE) using python 3.10 programming language
libraries and provides information through the monitor.
`
42
RESULT
Figure 10. Scenario 1. > 1m distance
Plate 1. Standing beside Plate 2. Standing behind
Plate 3. falling in line
`
43
Plate 5. Sitting with obstacle Plate 6. Standing sideways
Plate 7. Giving object to the other Plate 8. Standing with obstacle
`
44
Table 3
The researcher run a test in a system to get the result of accuracy in > 1 meter by using
confusion matrix in different position, by manually calculating the data gathered from the
system in
`
45
Figure 11 Scenario 2. <1m distance
Plate 1 . Handshake Test 2. Picking up an object
Plate 3. Hugging Plate 4. Reading
`
46
Plate 5. Standing and sitting Plate 6. Standing with obstacle
Plate 7. Sitting beside each other Plate 8. Sitting with obstacle
`
47
TABLE 4
The researcher run a test in a system to get the result of accuracy in <1 meter by using
confusion matrix in different position, by manually calculating the data gathered from the
system in Fig 10.
`
48
Table 5.
True Positive (TP): False Positive (FP):
 Reality: Person  Reality: Person
 Model predicted: <1m  Model predicted: >1m violate
violate social distance social distance
Number of FP results: 2
 Number of TP results: 4
False Negative (FN): True Negative (TN):
 Model predicted: <1m not  Model predicted: >1m not
detected detected
Number of FN results: 4 Number of TN results: 6
Confusion Matrix table.
Accuracy = _______TP + TN________ =

TP + TN + FP + FN
Accuracy = _______4 +6______ = 0.62 or 62.5%

4+6+2+4
Accuracy comes out to 0.62 or 62.5% ( 10 correct predictions out of 16 scenarios). Of the
16 scenarios, (6 TNs and 2 FPs) and (4 TPs and 4 FNs).
`
49
True Positives (TPs): 4 False Positives (FPs): 2
False Negatives (FNs): 4 True Negatives (TNs): 6

Precision = ______TP______
TP + FP
Precision = _______TP______ = ____4_ __ = 0.66 or 66%
TP + FP 4+2
Table 6
The model has a precision of 0.66 - in other words, when it detects the distance of a
person to the other, it is correct 66% of the time.
Recall = ___ TP___

TP + FN
Table 7
`
50
Recall = ____TP____ = ____4_ __ = 0.5 or 50%

The model has
TP a+recall
FN of 0.5 4- in
+ 6other words, it correctly detects 50% of all the Less
than 1-meter scenarios.
In this section shows the flow of gathering data relevant on the research study.
The researchers presented a formal letter of request to the school principal of C.V. Diez
Memorial Central Elementary School asking consent to conduct a study in a classroom
where limited face to face classes is implemented. Upon approval of the said request, the
researchers conducted an observation within the room by setting-up the computer and the
camera and with the use of the system the researcher can now gathered and analyze the
data.
`
51
Figure 12. Classroom where project conducted
`
52
Figure 13. Monitoring the social distance in the conducted area.
The classroom has a total of 16 person, 14 are students and 2 are teachers, to
observe the performance of the system and calculate the accuracy from the data gathered
using Confusion Matrix.
Table 8.
`
53
True Positive (TP): False Positive (FP):
 Model predicted: <1m  Model predicted: >1m
violate social distance violate social distance
Number of FP results: 5
 Number of TP results: 3
False Negative (FN): True Negative (TN):
 Model predicted: <1m not  Model predicted: >1m not
detected detected
Number of FN results: 0 Number of TN results: 8
Confusion Matrix
Accuracy = _______TP + TN________ =

TP + TN + FP + FN
Accuracy = _______3 +8______ = 0.78 or 78.5%

3+8+5+0
`
54
Accuracy comes out to 0.62 or 62.5% ( 11 correct predictions out of 16 total person). Of
the 16 scenarios, (8 TNs and 5 FPs) and (3 TPs and 0 FNs).
Precision = ______TP______
TP + FP
TABLE 9
Precision = _______TP______ = ____3_ __ = 37.5%

TP + FP 3+5
The model has a precision of 37.5% - in other words,, even two person is >1m distance
the other it is detected as a violation 37.5% at the time
Recall = ___ TP___

TP + FN
Table 10
Recall = ____TP____ = ____3_ __ = 100%

`
55
The model has a recall of 100% - in other words, people less than 1-meter will
detects 100% all the time.
`
56
CHAPTER 5
SUMMARY, FINDINGS,RECOMMENDATION AND CONCLUSION
This chapter presents the, conclusion derived from the findings and the
researcher’s recommendations and related to the implementation of the proposed project.
SUMMARY
The project Social Distancing Monitoring with Alert System for Limited face to
face class for Covid-19 was developed to help the school in this upcoming face to face
class. It can apply to any specific room in a school to observe and monitor social
distancing. Specifically, this system aimed to help the student by maintaining social
distancing whenever they see a camera in the area.
`
57
FINDINGS
1. The hardware input used was able to acquire atleast all of the features for running
and testing the system.
2. Social distance monitoring with alarm system was developed and operational;
also, the technical operation was properly documented.
3. The use of Python is a good environment for the system because it is user
friendly, easy to operate especially in creating system. The Python provides a
large library
features which can be operated by amateur and advanced users.
4. The overall accuracy of the system in detecting the social distance violation is
78.5%
RECOMMENDATION
After the conducting of the project, the researcher recommend:
 Higher specification of a computer to fast and more smoother
performance of the system
 Wide and bird view angle of camera is needed to achieve accurate
distance detection result.
`
58
 The next researcher will corporate the evaluation by the user in terms of
the system performance with the evaluation form given.
 For the future developers the system can be improved into prototype with
better performance and more features in the system .
Level of Agreement
1.) Very satisfactory (4)
2.) Satisfactory (3)
3.) Less Satisfactory (2)
4.) Not Satisfactory (1)
FUNCTIONALITY INDACATOR 4 3 2 1
Suitability The components of the system are
appropriate for the system needs
Accurateness The system can get accurate result for
the social distancing
Security The system is a free to use anywhere to
help monitor the people
RELIABILITY INDACATOR 4 3 2 1
Maturity The system reaches its full functionality
upon the implementation and after the
series of test conducted by the
researcher.
Operability The system will automatically monitor
and alert people to be physically
distance themselves from one another.
Learnability This system is so simple to manage
making its easily to undestand.
EFFICIENCY INDACATOR 4 3 2 1
Behavior The developed framework succesfully
distinguishes individuals who walk too
near and breach/violates social distance.
`
59
Stability For the longer time of usage, the system

has been stable.
Testability The system can be percieved by the
researchers that is has done undergone
series of test and test runs before
implementing for it shows quality and
efficient funcionality
CONCLUSION
Based on the findings of the study, the following conclusions were derived:
1. The fps (frame rate per second) will drop and the sytem slows if the camera
captures larger number of people violates social distance .
2. The system has an accuracy of 78.5%for social distancing.
3. The system cant detect people if they are in nearest distance in camera.
4. The system will automatically notify if people are not observing social distancing.
5. For better accuracy of the system, camera needs a wider range and perfect angle
to cover the whole area or room.
`
60
REFERENCE
Bibliography
[1] "World Health Organization," [Online]. Available: https://covid19.who.int/.
[2] "Commission on Higher Education," [Online]. Available:

https://ched.gov.ph/expansion-of-limited-face-to-face-classes-to-other-degree-
programs-approved-by-prrd-ched/.
[3] "National library of Medicine," [Online]. Available:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7889035/.
[4] "Wikipedia," [Online]. Available:

https://en.wikipedia.org/wiki/Computer_monitor.
[5] "Wikipedia," [Online]. Available: https://en.wikipedia.org/wiki/Social_distancing.
[6] "Wikipedia," [Online]. Available: https://en.wikipedia.org/wiki/Algorithm.
[7] "Wikipedia," [Online]. Available: https://en.wikipedia.org/wiki/System.
[8] "Lexico," [Online]. Available: https://www.lexico.com/definition/framework.
[9] "The pixel pantry," [Online]. Available:

https://www.thepixelpantry.com/about/#:~:text=A%20pixel%20is%20a
%20minute,and%20off%20our%20computer%20screens..
[10] "Computer vision," [Online]. Available:

https://sites.google.com/site/mechatronicssdl/research-groups/computer-vision.
[11] "LEXICO," [Online]. Available: https://www.lexico.com/definition/real_time.
[12] "Kullabs," [Online]. Available:

https://kullabs.com/class-miscellaneous/caomputer-basics/basic-concept/what-is-
a-computer#:~:text=A%20computer%20is%20an%20electronic,it%20in%20a
`
61
%20variable%20program..
[13] "Law Insider," [Online]. Available: https://www.lawinsider.com/dictionary/alarm-

system.
[14] "World Health Organization," [Online]. Available: https://www.who.int/news-

room/questions-and-answers/item/coronavirus-disease-covid-19-how-is-it-
transmitted.
[15] "Center For Disease Control and Prevention," 2000. [Online]. Available:
https://www.who.int/westernpacific/emergencies/covid-19/information/physical-
distancing.
[16] M. Rodriguez, I. Laptev, J. Sivic and J.-Y. Audibert, "Ieeexplore," 6-13 November
2011. [Online]. Available: https://ieeexplore.ieee.org/document/6126526.
[17] R. T. Collins, A. J. Lipton, T. Kanade, H. Fujiyoshi, D. Duggins, Y. Tsin, D.

Tolliver, N. Enomoto, O. Hasegawa, P. Burt and L. Wixson, "Citeseerx," june
2000. [Online]. Available: http://citeseerx.ist.psu.edu/viewdoc/summary?
doi=10.1.1.33.8388.
[18] T. XuanTung and Jong-MyonKim, "sciencedirect," july 2011. [Online]. Available:

https://www.sciencedirect.com/science/article/abs/pii/S0379711211000567.
[19] OnurKaraman, A. Alhudaif and KemalPolat, "Science direct," October 2021.

[Online]. Available:
https://www.sciencedirect.com/science/article/pii/S1568494621005317.
[20] "Medium," [Online]. Available: If they cross the acceptable social distance
threshold our camera will respond by sending the detected people to the System,
where our code resides. The code will wait for the moment when the social
distance threshold has been violated and immediately sou.
[21] "GeeksforGeeks - Confusion matrix Machine Learning," [Online]. Available:

https://www.geeksforgeeks.org/confusion-matrix-machine-learning/. [Accessed 13
December 2019].
`
62
`
63
APPENDICES A
Letter for Adviser
`
64
`
65
APPENDICES C
Letter to Conduct Study
`
66
APPENDICES D
`
67
Social Distance Detection
# import
the
necessar
y
packages
from .social_distancing_config import NMS_THRESH
from .social_distancing_config import MIN_CONF
import numpy as np
import cv2
def detect_people(frame, net, ln, personIdx=0):

# grab the dimensions of the frame and initialize the list of
# results
(H, W) = frame.shape[:2]
results = []
# construct a blob from the input frame and then perform a forward
# pass of the YOLO object detector, giving us our bounding boxes
# and associated probabilities
blob = cv2.dnn.blobFromImage(frame, 1 / 255.0, (416, 416),
swapRB=True, crop=False)
net.setInput(blob)
layerOutputs = net.forward(ln)
# initialize our lists of detected bounding boxes, centroids, and

# confidences, respectively
boxes = []
centroids = []
confidences = []
`
68
# loop over each of the layer outputs

for output in layerOutputs:
# loop over each of the detections
for detection in output:
# extract the class ID and confidence (i.e., probability)
# of the current object detection
scores = detection[5:]
classID = np.argmax(scores)
confidence = scores[classID]
# filter detections by (1) ensuring that the object

# detected was a person and (2) that the minimum
# confidence is met
if classID == personIdx and confidence >
MIN_CONF:
# scale the bounding box coordinates back
relative to
# the size of the image, keeping in mind that
YOLO
# actually returns the center (x, y)-coordinates
of
# the bounding box followed by the boxes'
width and
# height
box = detection[0:4] * np.array([W, H, W, H])
(centerX, centerY, width, height) =
box.astype("int")
# use the center (x, y)-coordinates to derive the

top
`
69
# and and left corner of the bounding box

x = int(centerX - (width / 2))
y = int(centerY - (height / 2))
# update our list of bounding box coordinates,

# centroids, and confidences
boxes.append([x, y, int(width), int(height)])
centroids.append((centerX, centerY))
confidences.append(float(confidence))
# apply non-maxima suppression to suppress weak, overlapping

# bounding boxes
idxs = cv2.dnn.NMSBoxes(boxes, confidences, MIN_CONF,
NMS_THRESH)
# ensure at least one detection exists

if len(idxs) > 0:
# loop over the indexes we are keeping
for i in idxs.flatten():
# extract the bounding box coordinates
(x, y) = (boxes[i][0], boxes[i][1])
(w, h) = (boxes[i][2], boxes[i][3])
# update our results list to consist of the person

# prediction probability, bounding box coordinates,
# and the centroid
r = (confidences[i], (x, y, x + w, y + h), centroids[i])
results.append(r)
`
70
# return the list of results

return results
Distance Detection
# base
path to
YOLO
directory
MODEL_PATH = "yolo-coco"
# initialize minimum probability to filter weak detections along with

# the threshold when applying non-maxima suppression
MIN_CONF = 0.3
NMS_THRESH = 0.3
# boolean indicating if NVIDIA CUDA GPU should be used

USE_GPU = False
# define the minimum safe distance (in pixels) that two people can be
# from each other
MIN_DISTANCE = 50
`
71
Social Distance Detection with Alarm
# import the necessary packages
from playsound import playsound
from tracker import social_distancing_config as config
from tracker.detection import detect_people
from scipy.spatial import distance as dist
import numpy as np
import argparse
import imutils
import cv2
import os
# construct the argument parse and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-i", "--input", type=str, default="",
help="path to (optional) input video file")
`
72
ap.add_argument("-o", "--output", type=str, default="",
help="path to (optional) output video file")
ap.add_argument("-d", "--display", type=int, default=1,
help="whether or not output frame should be displayed")
args = vars(ap.parse_args())
# load the COCO class labels our YOLO model was trained on
labelsPath = os.path.sep.join([config.MODEL_PATH, "coco.names"])
LABELS = open(labelsPath).read().strip().split("\n")
# derive the paths to the YOLO weights and model configuration
weightsPath = os.path.sep.join([config.MODEL_PATH, "yolov3.weights"])
configPath = os.path.sep.join([config.MODEL_PATH, "yolov3.cfg"])
# load our YOLO object detector trained on COCO dataset (80 classes)
print("[INFO] loading YOLO from disk...")
net = cv2.dnn.readNetFromDarknet(configPath, weightsPath)
`
73
# check if we are going to use GPU
if config.USE_GPU:
# set CUDA as the preferable backend and target
print("[INFO] setting preferable backend and target to CUDA...")
net.setPreferableBackend(cv2.dnn.DNN_BACKEND_CUDA)
net.setPreferableTarget(cv2.dnn.DNN_TARGET_CUDA)
# determine only the *output* layer names that we need from YOLO
ln = net.getLayerNames()
ln = [ln[i - 1] for i in net.getUnconnectedOutLayers()]
# initialize the video stream and pointer to output video file
print("[INFO] accessing video stream...")
vs = cv2.VideoCapture(0)
writer = None
`
74
# loop over the frames from the video stream
while True:
# read the next frame from the file
(grabbed, frame) = vs.read()
# if the frame was not grabbed, then we have reached the end
# of the stream
if not grabbed:
break
# resize the frame and then detect people (and only people) in it
frame = imutils.resize(frame, width=700)
results = detect_people(frame, net, ln,
personIdx=LABELS.index("person"))
# initialize the set of indexes that violate the minimum social
# distance
`
75
violate = set()
# ensure there are *at least* two people detections (required in
# order to compute our pairwise distance maps)
if len(results) >= 2:
# extract all centroids from the results and compute the
# Euclidean distances between all pairs of the centroids
centroids = np.array([r[2] for r in results])
D = dist.cdist(centroids, centroids, metric="euclidean")
# loop over the upper triangular of the distance matrix
for i in range(0, D.shape[0]):
for j in range(i + 1, D.shape[1]):
# check to see if the distance between any two
# centroid pairs is less than the configured number
# of pixels
if D[i, j] < config.MIN_DISTANCE:
`
76
# update our violation set with the indexes of
# the centroid pairs
violate.add(i)
violate.add(j)
# loop over the results
for (i, (prob, bbox, centroid)) in enumerate(results):
# extract the bounding box and centroid coordinates, then
# initialize the color of the annotation
(startX, startY, endX, endY) = bbox
(cX, cY) = centroid
color = (0, 255, 0)
# if the index pair exists within the violation set, then
# update the color
if i in violate:
color = (0, 0, 255)
`
77
playsound("beep.wav")
# draw (1) a bounding box around the person and (2) the
# centroid coordinates of the person,
cv2.rectangle(frame, (startX, startY), (endX, endY), color, 2)
cv2.circle(frame, (cX, cY), 5, color, 1)
# draw the total number of social distancing violations on the
# output frame
text = "Social Distancing Violations: {}".format(len(violate))
cv2.putText(frame, text, (10, frame.shape[0] - 25),
cv2.FONT_HERSHEY_SIMPLEX, 0.85, (0, 0, 255), 3)
# check to see if the output frame should be displayed to our
# screen
if args["display"] > 0:
# show the output frame
`
78
cv2.imshow("Frame", frame)
key = cv2.waitKey(1) & 0xFF
# if the `q` key was pressed, break from the loop
if key == ord("q"):
break
# if an output video file path has been supplied and the video
# writer has not been initialized, do so now
if args["output"] != "" and writer is None:
# initialize our video writer
fourcc = cv2.VideoWriter_fourcc(*"MJPG")
writer = cv2.VideoWriter(args["output"], fourcc, 25,
(frame.shape[1], frame.shape[0]), True)
# if the video writer is not None, write the frame to the output
# video file
`
79
if writer is not None:
writer.write(frame)
Source Code:
https://github.com/abd-shoumik/Social-distance-detection
`
80
CURRICULUM VITAE
PERSONAL DATA
Name : Conte, Mark Jason M.
Nickname : Mak mak
Date of Birth : August 14, 1998
Place of Birth : Surigao City
Home Address : Dapa, Surigao Del Norte
Civil Status : Single
Nationality : Filipino
Religion : SDA
Email – Address : markconte23@gmail.com
Parents : Mrs. Rimelyn M. Conte
Mr. Jorge P. Conte
EDUCATIONAL BACKGROUND
Tertiary : St. Paul University Surigao
Cor. San Nicolas & Rizal Streets, Surigao City
Elementary : Don Enrique Navarro Memorial School
Dapa, Surigao Del Norte
High School : San Nicolas School
Dapa, Surigao Del Norte
`
81
CURRICULUM VITAE
PERSONAL DATA
Name : Cillo, Francis Rae R.
Nickname : Rae
Date of Birth : May 1, 1998
Home Address : P-8 Navalca Barangay San Juan Surigao City
Nationality : Filipino
Religion : SDA
Email – Address : cillofrancis737@gmail.com
Parents : Mrs. Raema R. Cillo
Mr. Francisco S. Cillo
EDUCATIONAL BACKGROUND
Elementary : C.V. Diez Memorial Central Elementary School
Surigao City
High School : Surigao Norte National High School
Surigao City
`
82
CURRICULUM VITAE
Personal Information
Name : BESARIO, Drake Joshua J.

Date of Birth : July 10, 1998
Course : Bachelor of Science in Computer Engineering
Gender : Male
Parents : Deo R. Besario
Glenda J. Besario
Religion : Roman Catholic
Citizenship : Filipino
Home Address : M.Ortiz Street, Surigao City
E-mail Address : drakebesario3@gmail.com
Educational Background

Secondary : St. Paul University Surigao
Brgy. Luna, Surigao City, Surigao Del Norte
Elementary : St. Paul University Surigao
`
83

Social Distancing Monitoring With Alarm System For Limited Face To Face Classes in Covid 19

Uploaded by

Document Information

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

Social Distancing Monitoring With Alarm System For Limited Face To Face Classes in Covid 19

Uploaded by

Copyright:

Available Formats

St.

Paul University Surigao

SOCIAL DISTANCE MONITORING WITH ALARM SYSTEM FOR LIMITED

A Thesis Proposal Presented to

FRANCIS RAE R. CILLO

SOCIAL DISTANCE MONITORING WITH ALARM SYSTEM FOR LIMITED

Engr. Angelus Vincent P. Guilalas

Sr. Marie Rosanne B. Mallillin, SPC

ACCEPTED in partial fulfillment of the requirements for the degree of

___Engr. Angelus Vincent P. Guilalas___

material previously published or written by another person nor material which to a

have worked at Saint Paul University Surigao or elsewhere, is explicitly acknowledged in

FRANCIS RAE R. CILLO

MARK JASON M. CONTE

DRAKE JOSHUA J. BESARIO

assistance; To all of you, the researchers are thankful.

FRANCIS RAE R. CILLO

MARK JASON M. CONTE

DRAKE JOSHUA J. BESARIO

Social distancing throughout to be and adequate precaution against the spread of

contact among people.

can be implemented in camera by the use of Pycharm, an Integrated Development

50 (N pixel) and set a threshold. A violation threshold is established to identify whether

distance violation is 78.5%

Keywords: Computer vision, Person detection, Algorithm, Pycharm (IDE), Python

CERTIFICATE OF ORIGINALITY iii

THE PROBLEM AND ITS BACKGROUND 9

Statement of Purpose and Objectives 14

THE PROBLEM AND ITS BACKGROUND

COVID-19 originated from Wuhan; China has affected many countries

to public safety (World Health Organization, 2020). [1]

In the Philippines, according to the Department of Health (DOH), the statistics of

at all education levels.

Commission on Higher Education (CHED) proposed the limited face-to-face class to

the expansion proposal. [2]

to notify the people for violating the minimum distance.

This designed project is to help the institutions to minimized the crowd in a

project for the future.

Conceptual Framework of the Study

human detection in a video sequence and uses a triangular similarity technique to

measure the distance between persons detected by the camera.

The overall conceptual diagram is presented in work in Figure 1 is divided into

violations and further processed it to the system unit. [3]

Camera Input Frame

Detect violations and mark red PersonDetection

Speaker alarm output

Figure 1. Conceptual Diagram

Integrated Development Environment, and Python 3.10 as a programming language, as

INPUT PROCESS OUTPUT

Python 3.10 Development of System

Performance of the Design

Figure 2. Flow of the study

Statement of Purpose and Objective

This project is to help the institutions to observing social distancing. As a result of

system in social distancing monitoring with alarm.

conduct of limited face to face class

Development Environment (IDE) based on python program language.

3. To evaluate the performance of the design project based on different scenarios in

terms of precision, accuracy and sensitivity using confusion matrix.

where actual face-to-face class implemented.

• Accurate result of social distancing detection.

• Accessible components and resources.

• Achieve the monitoring social distance system with alarm.

_Engr. Angelus Vincent P. Guilalas_