“CHILD TRACKING DEVICE USING YOLOV3

TECHNOLOGY”

A Report on
Project Phase-II

Submitted in partial fulfillment of the requirement for the award of

Degree in Electrical & Electronics Engineering 8th semester of the

Visvesvaraya Technological University, Belagavi.

Submitted by
PROJECT ASSOCIATES USN
MANISH SAHANI 1SB17EE015
SUSHMA.N 1SB17EE029
NIKHIL.C.M 1SB17EE019
Under the guidance of
Dr.RAMANA REDDY
Asst. Prof., Dept. of EEE
Ms. MALINI K V
Asst. Prof. & HOD, Dept. of EEE

SRI SAIRAM COLLEGE OF ENGINEERING

Accredited by NAAC, ISO 9001:2015 Institution
SAI LEO NAGAR, GUDDANAHALLI (P.O), ANEKAL, BENGALURU-562106.

DEPARTMENT OF ELECTRICAL AND ELECTRONICS ENGINEERING

2020-2021
 SRI SAIRAM COLLEGE OF ENGINEERING
DEPARTMENT OF ELECTRICAL AND ELECTRONICS
ENGINEERING ANEKAL, BENGALURU-562106

CERTIFICATE
Certified that the project work entitled “CHILD TRACKING DEVICE
USING YOLOV3 TECHNOLOGY” a bonafide work carried out by
MANISH SAHANI (1SB17EE015), SUSHMA N (1SB17EE029),
NIKHIL C M (1SB17EE019) in partial fulfillment for the award of
Bachelor of Engineering in Electrical and Electronics Engineering of
the Visvesvaraya Technological University, Belagavi during the year
2020-2021. It is certified that all corrections/suggestions indicated
for internal assessment have been incorporated in the report
deposited in the departmental library. The project report has been
approved as it satisfied the academic requirement in respect of
project work prescribed for Bachelor of Engineering Degree.

Signature of Guide Signature of HOD Signature of Principal

............................ ........................... .............................

1. Dr. Ramana Reddy Ms. Malini K V Dr.B Shadaksharappa
Asst. Prof Asst. Prof. & HOD Principal
Dept. of EEE Dept. of EEE SSCE

..............................
2. Ms. Malini K V
Asst. Prof. & HOD,
Dept. of EEE

Sl. No Name of the Student USN

1 MANISH SAHANI 1SB17EE015
2 SUSHMA N 1SB17EE029
3 NIKHIL C M 1SB17EE019
External Viva
Name of the Examiners Signature with date
1) …………………………. …………………………
2) …………………………
…………………………
 ACKNOWLEDGEMENT
The successful completion of any project depends on help and co-
ordination of many people other than those who directly execute the
work.
It is difficult to express in words our profound sense of gratitude
to those who helped us but we make sense of gratitude to do so.

We wish to express our deep gratitude and indebtedness to our

institution SRI SAIRAM COLLEGE OF ENGINEERING, which has
provided us with the opportunity in fulfilling our desire of becoming an
Electrical and Electronics Engineers.
We would like to express our immense gratitude to our beloved
chairman & CEO Sri. SAI PRAKASH LEO MUTHU for their support for
the success of this project.
We are grateful to Dr. B. SHADAKSHARAPPA, Principal SSCE &
Ms. MALINI K V, Asst. Prof. & HOD of Electrical and Electronics
Engineering department, for their moral support and academic
guidance through our period of study and completion of this project.

We sincerely acknowledge our deep sense of gratitude for our

guide Dr. RAMANA REDDY, Asst. Prof. & Ms. MALINI K V, Asst. Prof.
& HOD, Dept. of EEE for their timely suggestions along with constant
evaluations of our progress which helped us to speed up the work and
make it highly knowledge oriented. Our special thanks to them.

Our own lab Technical Staffs have lent their hand whenever
necessary. Our special thanks to all technical staff’s.

We take this opportunity to thank various people assisted us

either directly or indirectly during execution and completion of our
project.
Last but not least, I wish to thank our Parents and all our friends
who extended their help and Co-operation.
 ABSTRACT

The main objective of the project is to design a child tracking system. The system is
developed to have a live track on the child with a quick alarming feature enabled as soon as it
gets apart from a fixed distance. The device will also be enabled with facial recognition. The
device will usually have two devices interlinked in which one is of parental or the officials.
But at the same time we are also link a network of devices so that the nearest device or the
authority will reach to the child. The technologies which we are planning at present to use are
GPS, Bluetooth module, YOLOV3 for facial recognition. This paper discusses the concept of
a smart wearable device for little children. The major advantage of this wearable over other
wearable is that it can be used in any cell phone and doesn't necessarily require an expensive
Smartphone and not a very tech savvy individual to operate. The purpose of this device is to
help parents locate their children with ease. At the moment there are many wearable’s in the
market which help track the daily activity of children and also help find the child using GPS
services present on the device. Therefore, the focus of this paper is to have an SMS text
enabled communication medium between the child's wearable and the parent as the
environment for GSM mobile communication is almost present everywhere.., the wearable
device will send with a text containing the real time accurate location of the child is 2m far
from home which upon tapping will provide directions to the child's location on Google maps
app, The prime motivation behind this paper is that we know how important technology is in
our lives but it can sometimes can't be trusted, and we always need to have a secondary
measure at hand.
 CONTENTS
SL.NO TOPICS PAGE NO

1 INTRODUCTION 7
1.1OBJECTIVE

2 LITERATURE SURVEY 8

3 METHODOLOGY 10
3.1 BLOCK DIAGRAM 10
3.2 WORK SEQUENCE FLOW

4 COMPONENTS USED 12

5 HARDWARE REQUIREMENTS 13
5.1 ARDUINO UNO
5.2 GSM MODULE
5.3 ULTRASONIC SENSOR
5.4 LCD DISPLAY
5.4.1 PIN DESCRIPTION
5.5 BUZZER
5.6 USB CAMERA

6 SOFTWARE REQUIREMENTS
6.1 AURDINO SOFTWARE
6.1.1 WRITING SKETCHES
6.1.2 FILE
6.1.3 EDIT
6.1.4 SKETCH
6.1.5 TOOLS
6.1.6 HELP
6.2 SOFTWARE REQUIREMENTS SPECIFICATIONS
6.3 PYTHON
6.3.1 PYTHON FEATURES
6.3.2 LIMITATIONS OF PYTHON
6.3.3 FLAVORS OF PYTHON
6.3.4 APPLICATIONS OF PYTHON
6.3.5 VERSIONS OF PYTHON
 6.3.6 PYTHON IN MACHINE LEARNING
6.4 WHAT IS MACHINE LEARNING
6.4.1 APPLICATIONS OF MACHINE LEARNING
ALGORITHMS
6.5 LIBRARIES AND PACKAGES
6.5.1 INSTALLATION
6.6 OPENCV- PYTHON
6.7 YOLO-ALGORITHM
6.7.1 WHAT IS YOLO OBJECT DETECTOR?
6.7.2 WHY THE YOLO ALGORITHM IS IMPORTANT?
6.8 YOLOV3 ALGORITHM
6.8.1 WHAT IS YOLOV3?
6.8.2 WHY THE NAME “YOU ONLY LOOK ONCE”?
6.8.3 HOW DOES YOLOV3 WORK?
6.8.4 THE ARCHITECTURE AT A GLANCE
6.8.5 WHATS NEW IN YOLOV3?
6.8.6 DISADVANTAGES OF YOLOV3 vs OTHER
ALGORITHMS
6.9 PREVIOUS METHODS
6.9.1 CLASSICAL FACE RECOGNITION ALGORITHM
6.9.2 ARTIFICIAL NEURAL NETWORKS IN FACE
RECOGNITION
6.9.3 GABOR WAVELET-BASED SOLUTION
6.9.4 3D-BASED FACE RECOGNITION
6.9.5 VEDIO-BASED FACE RECOGNITION
6.9.6 ADVANTAGES AND DISADVANTAGES
6.10 FACE RECOGNITION APPLICATIONS

7 SYSTEM ANALYSIS
7.1 FUNCTIONAL REQUIREMENTS
7.2 NON-FUNCTIONAL REQUIREMENTS

8 TEST RESULTS

9 ADVANTAGES
10 CONCLUSION
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 8

CHAPTER-1

INTRODUCTION
The motivation for this wearable comes from the increasing need for safety for little
children in current times as there could be scenarios of the child getting lost in the major
crowded areas. This paper focuses on the key aspect that lost child can be helped by the
people around the child and can play a significant role in the child's safety until reunited with
the parents. Most of the wearable’s available today are focused on providing the location,
activity, etc. of the child to the parents via GPS. Today, innovations is developing quickly and
giving all fundamental and successful answers for each prerequisite. Presently a day's
youngster security is a significant space of concern. This model is created to amend the
concerns of guardians in regards to their youngster security. In this situation, Our framework
guarantees greatest security and guarantees live following for their children since parent
stresses are certified. This paper proposed a model for kid security through advanced mobile
phones that gives the choice to follow the area of their kids just as in the event of crisis
youngsters can send a fast message and its present area by means of Short Message
administrations. This proposed framework is approved by testing on the Android stage.

1.1 OBJECTIVE

 The main objective of the project is to design a child tracking system.

• The system is developed to have a live track on the child with a quick alarming feature
enabled as soon as it gets apart from a fixed distance.
• The device will also be enabled with facial recognition.
• The device will usually have two devices interlinked in which one is of parental or the
officials. But at the same time we are also link a network of devices so that the nearest
device or the authority will reach to the child.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 9

CHAPTER-2

LITERATURE SURVEY

TITLE AUTHOR YEAR DESCRIPTION

Child Safety Aditi Gupta, Vibhor 2016 Today, technology is growing rapidly and
Harit-
& Tracking providing all essential and effective solutions
Department of
Management for every requirement. Now a day's child
Computer Science
System by security is an important area of concern. This
& Engineering,
Using GPS, model is developed to rectify the worries of
North India Institute
Geo-Fencing parents regarding their child security. In this
of Technology,
& Android scenario, Our system ensures maximum
Najibabad, UP,
Application: security and ensures live tracking for their
India
An Analysis kids because parent worries are genuine. This
paper proposed a model for child safety
through smart phones that provides the option
to track the location of their children as well
as in case of emergency children is able to
send a quick message and its current location
via Short Message services. This proposed
system is validated by testing on the Android
platform.

Easy to Wear Dhananjay Kumar, 2019 This paper presents an over simplistic design
Child Saurav, Ankur of an instrument which is targeted for
Guarding Yadav, Sharmila guarding little children against child theft and
Gadget missing from crowded places like shopping
malls or fairs or other potential crowded
places. The most remarkable feature of this
easy to wear device for little kids is its ability
to be controlled remotely by using a cell
phone and global system for mobile
communication (GSM).

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 10

Smart Mohamad Zikriya, 2019 As we know the present era is with equal
Gadget For Parmeshwar M G, rights, where in both men and women are
Women Shanmukayyar taking equal responsibility in their respective
Safety Using Math, Shraddha works. Hence women are giving equal
IoT Tankasali competition next to men in all fields, they are
assigned works in both the even and odd shift.
Every single day women and young girls from
all walks of life are being assaulted, molested,
and raped.

The Impact Sundus M 2020 Technology plays vital role in our daily lives.
of using These include electronic games, home
Gadgets on computers, handheld devices, and different
Children type of gadgets. Gadgets are popular in
children and as likely in elders. In this norm
we can’t keep our children out of this. This
paper presents the impact of gadgets on
children in positive and negative manner.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 11

CHAPTER-3
METHODOLOGY

To protect the safety of children we are making use of GPS tracking to pinpoint the
exact location of the child. This location link will be sent to parent’s smartphone through SMS
message service which can be used by Google maps application to track the child. Distance
between caretaker and children will be monitored continuously, when the child is separated
more than the safety limit a sound alert will be made to draw the attention of the caretaker
towards the child. Python YOLO is used to recognize the image captured by the device and
this image will be sent to parent through email service.

Ultrasonic sensor is used to measure the distance between child and parent. Distance in
cm is displayed on LCD screen and serial monitor. If distance is more than predetermined
safety limit then buzzer is turned ON to alert the parent. Message for child is moving far from
parent is shown on LCD display. GPS module is used send the latitude and longitude
positions to Arduino Uno. The GPS location data is displayed on the LCD screen. GPS
location data is sent to parent’s smartphone with location link through SMS message service
from GSM module. Google maps application is used to track the child location. Camera takes
the image of the person caring the child. This image is processed through python and sent to
parent through email service.

3.1 BLOCK DIAGRAM

ULTRASONIC BUZZE
SENSOR ARDUIO R
UNO
GSM

GPS
LCD
DISPLAY

Fig: Block diagram of system.

CAMER PYTHON EMAIL
Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
A YOLO
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 12

A Arduino controls the system architecture of the wearable with an Arduino Uno boot-
loader. A 5 pin header allows for power (+3 V) and ground connections as well as providing
access to TX, RX, and reset pins of the Arduino. The Fig illustrates the architecture of the
child safety wearable device, which depicts the various technologies and technological
standards used. The system architecture of the wearable is based and controlled by Arduino
with an Arduino Uno boot loader. The Arduino Uno collects various types of data from the
different modules interfaced to it, such as the GPS module upon being triggered by the
Arduino. The wearable device, for now, is not built on a SOC model, rather has been
proposed using larger components and can later build on the SOC platform once put into
manufacture. The wearable IOT device tasked with acquiring various data from the all the
different modules connected. The recorded time and identity will be sent through a message.
The wearable system runs on a battery with an output voltage of 5V. In order to maximize
power consumption, the wearable device has been programmed to provide GPS information
only upon request by SMS text via GSM shield.

3.2 WORK SEQUENCE FLOW

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 13

CHAPTER-4

COMPONENTS USED

Hardware Requirements:

 Arduino Uno.
 GSM Module.
 GPS Module.
 Buzzer.
 Ultrasonic Sensor.
 Camera.
 LCD Display.

Software Requirements:

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 14

 Arduino IDE.
 Python.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 15

CHAPTER-5

HARDWARE REQUIREMENTS

The most widely recognized arrangement of prerequisites characterized by any

working framework or programming application is the physical PC assets, otherwise
called equipment, an equipment necessities list is frequently joined by an equipment
similarity list, particularly if there should be an occurrence of working frameworks. A
HCL records tried, perfect, and now and then incongruent equipment gadgets for a
specific working framework or application. The accompanying sub-segments examine
the different parts of equipment prerequisites.

All PC working frameworks are intended for a specific PC design. Most

programming applications are restricted to specific working frameworks running on
specific structures. In spite of the fact that engineering free working frameworks and
applications exist, most should be recompiled to keep running on another design.

The energy of the focal preparing unit (CPU) is a central framework necessity for
any product. Most programming running on x86 engineering characterize preparing
power as the model and the clock speed of the CPU. Numerous different highlights of a
CPU that impact its speed and power, similar to transport speed, store, and MIPS are
frequently overlooked. This meaning of energy is regularly wrong, as AMD Intel
Pentium CPUs at comparative clock speed frequently have distinctive throughput
speeds.

• 10GB HDD(min)
• 128 MB RAM(min)
• Pentium P4 Processor 2.8Ghz(min)

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 16

5.1 Arduino UNO

Fig: Arduino UNO

Arduino is an open-source electronics platform based on easy-to-use hardware and
software. Arduino boards are able to read inputs - light on a sensor, a finger on a button, or a
Twitter message - and turn it into an output - activating a motor, turning on an LED,
publishing something online. You can tell your board what to do by sending a set of
instructions to the microcontroller on the board. To do so you use the Arduino programming
language (based on Wiring), and the Arduino Software (IDE), based on Processing.

Over the years Arduino has been the brain of thousands of projects, from everyday
objects to complex scientific instruments. A worldwide community of makers - students,
hobbyists, artists, programmers, and professionals - has gathered around this open-source
platform, their contributions have added up to an incredible amount of accessible
knowledge that can be of great help to novices and experts alike.

5.2 GSM Module

Fig: GSM Module

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 17

SIM800 is a quad-band GSM/GPRS module designed for the global market. It works
on frequencies GSM 850MHz, EGSM 900MHz, DCS 1800MHz and PCS 1900MHz. SIM800
features GPRS multi-slot class 12/ class 10 (optional) and supports the GPRS coding schemes
CS-1, CS-2, CS-3 and CS-4. With a tiny configuration of 24*24*3mm, SIM800 can meet
almost all the space requirements in users’ applications, such as M2M, smart phone, PDA and
other mobile devices. SIM800 has 68 SMT pads, and provides all hardware interfaces
between the module and customers’ boards. SIM800 is designed with power saving technique
so that the current consumption is as low as 1.2mA in sleep mode. SIM800 integrates TCP/IP
protocol and extended TCP/IP AT commands which are very useful for data transfer
applications.

5.3 Ultrasonic Sensor

Fig: Ultrasonic Sensor

Ultrasonic ranging module HC - SR04 provides 2cm - 400cm non-contact

measurement function, the ranging accuracy can reach to 3mm. The modules includes
ultrasonic transmitters, receiver and control circuit

 5V Supply
 Trigger Pulse Input
 Echo Pulse Output
 0V Ground

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 18

5.4 LCD Display

LCD (Liquid Crystal Display) screen is an electronic display module and find a wide
range of applications. A 16x2 LCD display is very basic module and is very commonly used
in various devices and circuits. These modules are preferred over seven segments and other
multi segment LEDs. The reasons being: LCDs are economical; easily programmable; have
no limitation of displaying special & even custom characters (unlike in seven segments)
animations and so on.

A 16x2 LCD means it can display 16 characters per line and there are 2 such lines. In
this LCD each character is displayed in 5x7 pixel matrix. This LCD has two registers, namely,
Command and Data.

The command register stores te command instructions given to the LCD. A command
is an instruction given to LCD to do a predefined task like initializing it, clearing its screen,
setting the cursor position, controlling display etc. The data register stores the data to be
displayed on the LCD. The data is the ASCII value of the character to be displayed on the
LCD. Click to learn more about internal structure of a LCD.

Fig: LCD Display

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 19

Fig: LCD Pin Description

5.4.1 Pin Description: 

 Pin
 Function  Name
No
1 Ground (0V) Ground
2 Supply voltage; 5V (4.7V – 5.3V)  Vcc
3 Contrast adjustment; through a variable resistor  VEE
4 Selects command register when low; and data register when high Register Select
5 Low to write to the register; High to read from the register Read/write
6 Sends data to data pins when a high to low pulse is given Enable
7 DB0
8 DB1
9 DB2
10 DB3
8-bit data pins
11 DB4
12 DB5
13 DB6
14 DB7
15 Backlight VCC (5V) Led+
16 Backlight Ground (0V) Led-
 

5.5 Buzzer

Fig. Buzzer.
The buzzer is a sounding device that can convert audio signals into sound signals. It is
usually powered by DC voltage. It is widely used in alarms, computers, printers and other
electronic products as sound devices. A buzzer or beeper is an audio signaling device, which
may be mechanical, electromechanical, or piezoelectric (Piezo for short). Typical uses
of buzzers and beepers include alarm devices, timers, and confirmation of user input such as a
mouse click or keystroke.
Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 20

5.6 USB Camera

Fig: USB Camera

 Up & down 30 degrees rotatable, you can adjust the angle as you like.
 Support windows 2000/ XP/ Win7/ Win8/ Win10/ Vista 32bit./ Mac. Designed for
both laptop and desktop, auto White balance, auto color correction. Support various video
meeting software, ie, net meeting and works great with msn, WeChat, QQ, Yahoo and
Skype etc.
 High definition and true color images, manual focus. Built-in sound absorption
Microphone, your voice can be heard clearly in 30 feet, meaning that you don't have to get
close to even kiss Your camera awkwardly.
 Imported Optical lens, High precision and no distorted pictures, compatible with USB.
 The computer would automatically install the driver in the Win7 and lower level
system after inserting this webcam, but in Win10, you don't need to install any driver, you
can directly use the video chat software features and then can see the picture.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 21

CHAPTER-6

SOFTWARE REQUIREMENTS

6.1 ARDUINO SOFTWARE (IDE)

The Arduino Integrated Development Environment - or Arduino Software (IDE) -

contains a text editor for writing code, a message area, a text console, a toolbar with buttons
for common functions and a series of menus. It connects to the Arduino and Genuino
hardware to upload programs and communicate with them.

Fig: Arduino Tool

6.1.1 Writing Sketches

Programs composed utilizing Arduino Software (IDE) are called portrays. These
representations are composed in the content tool and are spared with the document expansion
.ino. The proofreader has highlights for cutting/gluing and for looking/supplanting content.
The message zone gives criticism while sparing and sending out and furthermore shows
blunders. The comfort shows content yield by the Arduino Software (IDE), including
complete blunder messages and other data. The base right hand corner of the window shows
the arranged board and serial port. The toolbar catches permit you to check and transfer
programs, make, open, and spare outlines, and open the serial screen.

Upload 
Accumulates your code and transfers it to the designed board. See

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 22

transferring beneath for points of interest.

Note: If you are utilizing an outer software engineer with your

board, you can hold down the "move" key on your PC when
utilizing this symbol. The content will change to "Transfer utilizing
Programmer"

New 
Creates a new sketch.

Open 
Presents a menu of all the representations in your sketchbook.
Clicking one will open it inside the present window overwriting its
substance.

Note: because of a bug in Java, this menu doesn't scroll; on the off
chance that you have to open a draw late in the rundown, utilize the
File | Sketchbook menu.

Save 
Saves your sketch.

Serial Monitor 
Opens the serial monitor.

Additional commands are found within the five menus

o File
o Edit
o Sketch
o Tools
o Help.

The menus are setting delicate, which implies just those things significant to the work as of
now being done are accessible.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 23

Fig: Arduino Software

6.1.2 FILE

 New
Makes another occasion of the editorial manager, with the absolute minimum structure
of an outline already set up.

 Open 
Permits to stack an outline document perusing through the PC drives and
envelopes.

 Open Recent 
Provides a short list of the most recent sketches, ready to be opened.

 Sketchbook 
Demonstrates the current portrays inside the sketchbook envelope structure;

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 24

tapping on any name opens the relating sketch in another proofreader

example.

 Examples 
Any case gave by the Arduino Software (IDE) or library appears in this menu
thing. Every one of the cases are organized in a tree that permits simple access by
topic or library.

 Close 
Closes the instance of the Arduino Software from which it is clicked.

 Save 
Spares the draw with the present name. On the off chance that the record
hasn't been named some time recently, a name will be given in a "Spare as."
window.

 Save as 
Allows to save the current sketch with a different name.

 Page Setup 
It shows the Page Setup window for printing.

 Print 
Sends the current sketch to the printer according to the settings defined in
Page Setup.

 Preferences 
Opens the Preferences window where some settings of the IDE may be
customized, as the language of the IDE interface.

 Quit 
Shuts all IDE windows. The same representations open when Quit was picked
will be automatically revived whenever you begin the IDE.

6.1.3 EDIT

 Undo/Redo 
Backpedals of at least one stages you did while altering; when you backpedal,
you may go ahead with Redo.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 25

 Cut 
Expels the chose content from the supervisor and spots it into the clipboard.

 Copy 
Copies the chose message in the manager and spots it into the clipboard.

 Copy for Forum 

Duplicates the code of your portray to the clipboard in a frame appropriate for
presenting on the discussion, finish with sentence structure shading.

 Copy as HTML 
Duplicates the code of your draw to the clipboard as HTML, reasonable for
Implanting in website pages.

 Paste 
Puts the substance of the clipboard at the cursor position, in the editorial
manager.

 Select All 
Chooses and highlights the entire substance of the editorial manager.

 Comment/Uncomment 
Puts or evacuates the/remark marker toward the start of each chose line.

 Increase/Decrease Indent 
Includes or subtracts a space toward the start of each chose line, moving the
text one space on the privilege or dispensing with a space toward the start.

 Find 
Opens the Find and Replace window where you can indicate content to seek
inside the current draw as indicated by a few choices.

 Find Next 
Highlights the following event - if any - of the string indicated as the pursuit
thing in the Find window, with respect to the cursor position.

 Find Previous 
Highlights the past event - if any - of the string indicated as the hunt item in the
Find window with respect to the cursor position.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 26

6.1.4 SKETCH

 Verify/Compile 
Checks your outline for blunders arranging it; it will report memory utilization
for code and factors in the reassure territory.

 Upload 
Arranges and stacks the twofold document onto the designed board through
the configured Port.

 Upload Using Programmer 

This will overwrite the boot loader on the board; you should utilize Tools >
Burn Boot loader to reestablish it and have the capacity to Upload to USB
serial port once more. In any case, it permits you to utilize the full
limit of the Flash memory for your sketch. It would be ideal if you take note of that
this summon won't consume the wires. To do as such a Tools - >Burn Boot loader
charge must be executed.

 Export Compiled Binary 

Spares a .hex document that might be kept as chronicle or sent to the board
utilizing different tools.

 Show Sketch Folder 

Opens the current sketch folder.

 Include Library 
Adds a library to your portray by embeddings #include explanations toward the
begin of your code. For more points of interest, see libraries underneath. Furthermore,
from this menu thing you can get to the Library Manager and import new libraries
from .compress records.

 Add File... 
Adds a source record to the outline (it will be duplicated from its present area).
The new document shows up in another tab in the portray window. Documents
can be expelled from the portray utilizing the tab menu available tapping on the little
triangle symbol beneath the serial screen one on the correct side to the toolbar.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 27

6.1.5 TOOLS

 Auto Format 
This organizations your code pleasantly: i.e. indents it so that opening and
shutting wavy braces line down, and that the announcements inside wavy
supports are indented more.

 Archive Sketch 
Documents a duplicate of the present draw in .zip arrange. The file is set in the
same index as the draw.

 Fix Encoding & Reload 

Fixes conceivable errors between the supervisor burn outline and other
operating frameworks scorch maps.

 Serial Monitor 
Opens the serial screen window and starts the trading of information with any
connected board on the present chose Port. This as a rule resets the board, if
the board bolsters Reset over serial port opening.

 Board 
Select the board that you're utilizing. See beneath for depictions of the
different boards.

 Port 
This menu contains all the serial gadgets (genuine or virtual) on your machine.
It should consequently revive each time you open the top-level instruments
menu.

 Programmer 
For choosing a hardware software engineer when programming a board or chip
and not using the locally available USB-serial association. Typically you won't
require this, however in the event that you're consuming a bootloader to
another microcontroller, you will utilize this.

 Burn Boot loader 

The things in this menu permit you to consume a bootloader onto the
microcontroller on an Arduino board. This is not required for ordinary utilization of an

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 28

Arduino or Genuino board however is valuable in the event that you buy another
ATmega microcontroller Ensure that you've chosen the right board from the Boards
menu before consuming the bootloader on the objective board. This charge
additionally set the correct circuits.

6.1.6 HELP

Here you find simple access to various reports that accompany the Arduino Software (IDE).
You have entry to Getting Started, Reference, this manual for the IDE and different archives
locally, without a web association. The reports are a nearby duplicate of the online ones and
may connect back to our online site.

 Find in Reference 

This is the main intuitive capacity of the Help menu: it specifically chooses
the pertinent page in the neighborhood duplicate of the Reference for the
capacity or charge under the cursor.

6.2 Software Requirements Specifications

H/W System Configuration:

Device Arduino

S/W System Configuration:

Operating System Windows 10.

Technology Machine Learning.
IDLE Python 3.7 or higher.

Programming necessities manage characterizing programming asset necessities

and requirements that should be introduced on a PC to give ideal working of an
application.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
 CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 29

These necessities or requirements are for the most part excluded in the product
establishment bundle and should be introduced independently before the product is
introduced.
 Python 3.7 or higher
 Pycharm
 opencv

Outline of advances
The innovations utilized is depicted as underneath:

6.3 PYTHON

• Python is a general purpose high level programming Language (human

understandable languages are High level programming languages)
• Python Developed by Guido Van Rossam
• 1989 National Research Institute(NRI) At Netherland
• Officially Python available to the public in 1991 :: FEB 20th 1991
Python was imagined in the late 1980s, and its usage started in December 1989 by
Guido van Rossum at Centrum Wiskunde and Informatica (CWI) in the Netherlands as a
successor to the ABC dialect (itself roused by SETL) capable of exemption dealing with
and interfacing with the Amoeba working system. Van Rossum remains Python's chief
creator. His proceeding with focal part in Python's advancement is reflected in the title
given him by the Python people group:

6.3.1 Python Feature

• Simple and easy to learn

Python has only 33 keywords But JAVA has (53) keywords
• Free ware (There is no license we cannot pay anything)
& Open source (we can able to see source code if source is not good I can able to customize our
requirements)
• High level programming language (human understandable language)
Python Is Platform Independent (It means I can write a program once and run any where (WORA)
• Portability

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 30

Moving python program from one platform to another platform without changing any
thing.
• Dynamically Typed Programming Language
In python we are not required to declare type in Python
• Both Object Oriented and Procedure Oriented Language
• Interpreted Language
It means we are not going to compile
• Extensible
We can use Other Programming Language in Python

6.3.2 Limitations of python

• Performance wise it is not up to the mark because it is an interpreted language

Interpreter able to see only one line (JAVA is better performance compare to python in
java JIT (just in time compiler) concept is there mobile applications it is not up to the
mark.
Myth:-python is not suitable large scale enterprise applications

6.3.3 Flavors of python

• Cpython :- It can be standard, It can be used to c language python

• Jpython or jpython :- it is for JAVA application
• Iron python:-to work with microsoft .net platform
• Pypy :-Internally JIT (just intime compiler) compiler is there so performance wise too
good
• Ruby python:- used for ruby application
• Anaconda python:- To handle Big-data happily go for Anaconda python
• Stackless (python for concurrancy) :-
parallely you execute (like mutithread) so go for stackless.

6.3.4 APPLICATIONS OF PYTHON

 GUI-Based Desktop Applications:

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 31

Python has simple syntax, modular architecture, rich text processing tools and the
ability to work on multiple operating systems which make it a desirable choice for developing
desktop-based applications. There are various GUI toolkits like wxPython, PyQt or PyGtk
available which help developers create highly functional Graphical User Interface (GUI). The
various applications developed using Python includes:

 Image Processing and Graphic Design Applications:

Python has been used to make 2D imaging software such as Inkscape, GIMP,
Paint Shop Pro and Scribus. Further, 3D animation packages, like Blender, 3ds
Max, Cinema 4D, Houdini, Lightwave and Maya, also use Python in variable
proportions.

 Scientific and Computational Applications:

The higher speeds, productivity and availability of tools, such as Scientific Python
and Numeric Python, have resulted in Python becoming an integral part of
applications involved in computation and processing of scientific data. 3D
modeling software, such as FreeCAD, and finite element method software, such as
Abaqus, are coded in Python.

 Games:

Python has various modules, libraries and platforms that support development of
games. For example, PySoy is a 3D game engine supporting Python 3, and
PyGame provides functionality and a library for game development. There have
been numerous games built using Python including Civilization-IV, Disney’s
Toontown Online, Vega Strike etc.

 Web Frameworks and Web Applications:

Python has been used to create a variety of web-frameworks including CherryPy,

Django, TurboGears, Bottle, Flask etc. These frameworks provide standard libraries and
modules which simplify tasks related to content management, interaction with database and
interfacing with different internet protocols such as HTTP, SMTP, XML-RPC, FTP and POP.
Plone, a content management system; ERP5, an open source ERP which is used in aerospace,
Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 32

apparel and banking; Odoo – a consolidated suite of business applications; and Google App
engine are a few of the popular web applications based on Python.

 Enterprise and Business Applications:

With features that include special libraries, extensibility, scalability and easily
readable syntax, Python is a suitable coding language for customizing larger applications.
Reddit, which was originally written in Common Lips, was rewritten in Python in 2005.
Python also contributed in a large part to functionality in YouTube.

 Operating Systems:

Python is often an integral part of Linux distributions. For instance, Ubuntu’s

Ubiquity Installer, and Fedora’s and Red Hat Enterprise Linux’s Anaconda Installer are
written in Python. Gentoo Linux makes use of Python for Portage, its package management
system.

 Language Development:

Python’s design and module architecture has influenced development of numerous

languages. Boo language uses an object model, syntax and indentation, similar to Python.
Further, syntax of languages like Apple’s Swift, CoffeeScript, Cobra, and OCaml all share
similarity with Python.

 Prototyping:

Besides being quick and easy to learn, Python also has the open source advantage of
being free with the support of a large community. This makes it the preferred choice for
prototype development. Further, the agility, extensibility and scalability and ease of
refactoring code associated with Python allow faster development from initial prototype.Since
its origin in 1989, Python has grown to become part of a plethora of web-based, desktop-
based, graphic design, scientific, and computational applications. With Python available for
Windows, Mac OS X and Linux / UNIX, it offers ease of development for enterprises.
Additionally, the latest release Python 3.4.3 builds on the existing strengths of the language,
with drastic improvement in Unicode support, among other new features.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 33

6.3.5 Versions of python

• Python 1.0 Introduced in jan 1994

• Python 2.0 Introduced in oct 2000
• Python 3.0 introduced in dec 2008
latest version
python 3.6.3 è 2016
python 3.7
Any new version should provide support for old version programs
• There is no- backward compatibility support
• Python 3 is not support to python 2 program

6.3.6 Python in Machine Learning

Python has libraries that enables developers to use optimized algorithms. It implements
popular machine learning techniques such as recommendation, classification, and clustering.
Therefore, it is necessary to have a brief introduction to machine learning before we move
further.

6.4 What is Machine Learning?

Data science, machine learning and artificial intelligence are some of the top trending
topics in the tech world today. Data mining and Bayesian analysis are trending and this is
adding the demand for machine learning. This tutorial is your entry into the world of machine
learning.
Machine learning is a discipline that deals with programming the systems so as to make them
automatically learn and improve with experience. Here, learning implies recognizing and
understanding the input data and taking informed decisions based on the supplied data. It is
very difficult to consider all the decisions based on all possible inputs. To solve this problem,
algorithms are developed that build knowledge from a specific data and past experience by
applying the principles of statistical science, probability, logic, mathematical optimization,
reinforcement learning, and control theory.

6.4.1 Applications of Machine Learning Algorithms

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 34

The developed machine learning algorithms are used in various applications such as:
 Vision processing

 Language processing

 Forecasting things like stock market trends, weather

 Pattern recognition

 Games

 Data mining

 Expert systems

 Robotics

6.5 Libraries and Packages

To understand machine learning, you need to have basic knowledge of Python

programming. In addition, there are a number of libraries and packages generally used in
performing various machine learning tasks as listed below:
 numpy - is used for its N-dimensional array objects

 pandas – is a data analysis library that includes data frames

 matplotlib – is 2D plotting library for creating graphs and plots

 scikit-learn - the algorithms used for data analysis and data mining tasks

 seaborn – a data visualization library based on matplotlib

6.5.1 Installation

You can install software for machine learning in any of the two methods as discussed here:

 Method 1
Download and install Python separately from python.org on various operating systems as
explained below:
To install Python after downloading, double click the .exe (for Windows) or .pkg (for Mac)
file and follow the instructions on the screen.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 35

For Linux OS, check if Python is already installed by using the following command at the
prompt:
$ python --version. ...
If Python 2.7 or later is not installed, install Python with the distribution's package manager.
Note that the command and package name varies.
On Debian derivatives such as Ubuntu, you can use apt:
$ sudo apt-get install python3
Now, open the command prompt and run the following command to verify that Python is
installed correctly:
$ python3 --version
Python 3.6.2

Python Machine Learning – Environment Setup

Similarly, we can download and install necessary libraries like numpy, matplotlib etc.
individually using installers like pip. For this purpose, you can use the commands shown
here:
$pip install numpy
$pip install matplotlib
$pip install pandas
$pip install seaborn

 Method 2

Alternatively, to install Python and other scientific computing and machine learning
packages simultaneously, we should install Anaconda distribution. It is a Python
implementation for Linux, Windows and OSX, and comprises various machine learning
packages like numpy, scikit-learn, and matplotlib. It also includes Jupyter Notebook, an
interactive Python environment. We can install Python 2.7 or any 3.x version as per our
requirement.
To download the free Anaconda Python distribution from Continuum Analytics, you can do
the following:

Visit the official site of Continuum Analytics and its download page. Note that the
installation process may take 15-20 minutes as the installer contains Python, associated

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 36

packages, a code editor, and some other files. Depending on your operating system, choose
the installation process as explained here:

For Windows: Select the Anaconda for Windows section and look in the column with
Python 2.7 or 3.x. You can find that there are two versions of the installer, one for 32-bit
Windows, and one for 64-bit Windows. Choose the relevant one.

For Mac OS: Scroll to the Anaconda for OS X section. Look in the column with Python 2.7
or 3.x. Note that here there is only one version of the installer: the 64-bit version.
For Linux OS: We select the "Anaconda for Linux" section. Look in the column with Python
2.7 or 3.x.
Note that you have to ensure that Anaconda’s Python distribution installs into a single
directory, and does not affect other Python installations, if any, on your system.
To work with graphs and plots, we will need these Python library packages: matplotlib and
seaborn.
If you are using Anaconda Python, your system already has numpy, matplotlib, pandas,
seaborn, etc. installed. We start the Anaconda Navigator to access either Jupyter Note book or
Spyder IDE of python.
After opening either of them, type the following commands:
import numpy
import matplotlib Python Machine Learning
Now, we need to check if installation is successful. For this, go to the command line and type
in the following command:
$ python
Python 3.6.3 |Anaconda custom (32-bit)| (default, Oct 13 2017, 14:21:34)
[GCC 7.2.0] on linux
Next, you can import the required libraries and print their versions as shown:
>>>import numpy
>>>print numpy.__version__
1.14.2
>>> import matplotlib
>>> print (matplotlib.__version__)
2.1.2
>> import pandas
>>> print (pandas.__version__)
0.22.0
Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 37

>>> import seaborn

>>> print (seaborn.__version__)

6.6 OpenCV-Python

OpenCV-Python is a library of Python bindings designed to solve computer vision

problems.

Python is a general purpose programming language started by Guido van Rossum that

became very popular very quickly, mainly because of its simplicity and code readability. It
enables the programmer to express ideas in fewer lines of code without reducing readability.

Compared to languages like C/C++, Python is slower. That said, Python can be easily
extended with C/C++, which allows us to write computationally intensive code in C/C++ and
create Python wrappers that can be used as Python modules. This gives us two advantages:
first, the code is as fast as the original C/C++ code (since it is the actual C++ code working in
background) and second, it easier to code in Python than C/C++. OpenCV-Python is a Python
wrapper for the original OpenCV C++ implementation.

OpenCV-Python makes use of Numpy, which is a highly optimized library for numerical
operations with a MATLAB-style syntax. All the OpenCV array structures are converted to
and from Numpy arrays. This also makes it easier to integrate with other libraries that use
Numpy such as SciPy and Matplotlib.

6.7 YOLO algorithm

YOLO is an abbreviation for the term ‘You Only Look Once’. This is an algorithm that
detects and recognizes various objects in a picture (in real-time). Object detection in YOLO is
done as a regression problem and provides the class probabilities of the detected images.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 38

YOLO algorithm employs convolutional neural networks (CNN) to detect objects in

real-time. As the name suggests, the algorithm requires only a single forward propagation
through a neural network to detect objects.

This means that prediction in the entire image is done in a single algorithm run. The
CNN is used to predict various class probabilities and bounding boxes simultaneously.

The YOLO algorithm consists of various variants. Some of the common ones include
tiny YOLO and YOLOv3.

6.7.1 What is the YOLO object detector?

Figure 1: A simplified illustration of the YOLO object detector pipeline (source). We’ll use
YOLO with OpenCV in this blog post.
When it comes to deep learning-based object detection, there are three primary object
detectors you’ll encounter:

 R-CNN and their variants, including the original R-CNN, Fast R- CNN, and Faster R-
CNN

 Single Shot Detector (SSDs)

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 39

 YOLO

6.7.2 Why the YOLO algorithm is important?

YOLO algorithm is important because of the following reasons:

 Speed: This algorithm improves the speed of detection because it can predict objects
in real-time.
 High accuracy: YOLO is a predictive technique that provides accurate results with
minimal background errors.
 Learning capabilities: The algorithm has excellent learning capabilities that enable it
to learn the representations of objects and apply them in object detection.

6.8 YOLOV3 ALGORITHM

6.8.1 What is YOLOv3?

YOLOv3 (You Only Look Once, Version 3) is a real-time object detection algorithm that
identifies specific objects in videos, live feeds, or images. YOLO uses features learned by
a deep convolutional neural network to detect an object. Versions 1-3 of YOLO were created
by Joseph Redmon and Ali Farhadi.

The first version of YOLO was created in 2016, and version 3, which is discussed
extensively in this article, was made two years later in 2018. YOLO is implemented using the
Keras or OpenCV deep learning libraries.

Object classification systems are used by Artificial Intelligence (AI) programs to perceive
specific objects in a class as subjects of interest. The systems sort objects in images into
groups where objects with similar characteristics are placed together, while others are
neglected unless programmed to do otherwise.

6.8.2 Why the name “you only look once”?

As typical for object detectors, the features learned by the convolutional layers are
passed onto a classifier which makes the detection prediction. In YOLO, the prediction is
based on a convolutional layer that uses 1×1 convolutions.

YOLO is named “you only look once” because its prediction uses 1×1 convolutions; the
size of the prediction map is exactly the size of the feature map before it.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 40

Fig: YOLOV3 Computer Version

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 41

6.8.3 How does YOLOv3 work?

YOLO is a Convolutional Neural Network (CNN) for performing object detection in real-

time. CNNs are classifier-based systems that can process input images as structured arrays of
data and identify patterns between them. YOLO has the advantage of being much faster than
other networks and still maintains accuracy.

Fig: YOLOV3 Working

It allows the model to look at the whole image at test time, so its predictions are
informed by the global context in the image. YOLO and other convolutional neural network
algorithms “score” regions based on their similarities to predefined classes.

High-scoring regions are noted as positive detections of whatever class they most closely
identify with. For example, in a live feed of traffic, YOLO can be used to detect different
kinds of vehicles depending on which regions of the video score highly in comparison to
predefined classes of vehicles.

6.8.4 The Architecture at a Glance

The YOLOv3 algorithm first separates an image into a grid. Each grid cell predicts some
number of boundary boxes (sometimes referred to as anchor boxes) around objects that score
highly with the aforementioned predefined classes.

Each boundary box has a respective confidence score of how accurate it assumes that
prediction should be and detects only one object per bounding box. The boundary boxes are
generated by clustering the dimensions of the ground truth boxes from the original dataset to
find the most common shapes and sizes.
Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 42

Other comparable algorithms that can carry out the same objective are R-CNN (Region-
based Convolutional Neural Networks made in 2015) and Fast R-CNN (R-CNN improvement
developed in 2017), and Mask R-CNN.

However, unlike systems like R-CNN and Fast R-CNN, YOLO is trained to do
classification and bounding box regression at the same time.

6.8.5 What’s New in YOLOv3?

There are major differences between YOLOv3 and older versions occur in terms of speed,
precision, and specificity of classes. YOLOv2 and YOLOv3 are worlds apart in terms of
accuracy, speed, and architecture. YOLOv2 came out in 2016, two years before YOLOv3.

The following sections will give an overview of what’s new in YOLOv3.

 Speed
YOLOv2 was using Darknet-19 as its backbone feature extractor, while YOLOv3 now
uses Darknet-53. Darknet-53 is a backbone also made by the YOLO creators Joseph Redmon
and Ali Farhadi.

Darknet-53 has 53 convolutional layers instead of the previous 19, making it more
powerful than Darknet-19 and more efficient than competing backbones (ResNet-101 or
ResNet-152).

Comparison of backbones. Accuracy, billions of operations (Ops), billion floating-point

operations per second (BFLOP/s), and frames per second (FPS) for various networks

Using the chart provided in the YOLOv3 paper by Redmon and Farhadi, we can see
that Darknet-52 is 1.5 times faster than ResNet101. The depicted accuracy doesn’t entail any
trade-off between accuracy and speed between Darknet backbones either since it is still as
accurate as ResNet-152 yet two times faster.
Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 43

YOLOv3 is fast and accurate in terms of mean average precision (mAP) and intersection
over union (IOU) values as well. It runs significantly faster than other detection methods with
comparable performance (hence the name – You only look once).

Moreover, you can easily trade-off between speed and accuracy simply by changing the
model’s size, and no retraining required.

Fig: YOLOv3 runs much faster than other detection methods with a comparable
performance using an M40/Titan X GPU

 Precision for Small Objects

The chart below (taken and modified from the YOLOv3 paper) shows the average
precision (AP) of detecting small, medium, and large images with various algorithms and
backbones. The higher the AP, the more accurate it is for that variable.

The precision for small objects in YOLOv2 was incomparable to other algorithms
because of how inaccurate YOLO was at detecting small objects. With an AP of 5.0, it paled
compared to other algorithms like RetinaNet (21.8) or SSD513 (10.2), which had the second-
lowest AP for small objects.

YOLOv3 increased the AP for small objects by 13.3, which is a massive advance from
YOLOv2. However, the average precision (AP) for all objects (small, medium, large) is still
less than RetinaNet

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 44

YOLOv3 comparison for different object sizes showing the average precision (AP) for
AP-S (small object size), AP-M (medium object size), AP-L (large object size)

 Specificity of Classes
The new YOLOv3 uses independent logistic classifiers and binary cross-entropy loss for
the class predictions during training. These edits make it possible to use complex datasets
such as Microsoft’s Open Images Dataset (OID) for YOLOv3 model training. OID contains
dozens of overlapping labels, such as “man” and “person” for images in the dataset.

YOLOv3 uses a multilabel approach which allows classes to be more specific and be
multiple for individual bounding boxes. Meanwhile, YOLOv2 used a softmax, which is a
mathematical function that converts a vector of numbers into a vector of probabilities, where
the probabilities of each value are proportional to the relative scale of each value in the vector.

Using a softmax makes it so that each bounding box can only belong to one class, which
is sometimes not the case, especially with datasets like OID.

Fig: Object Detection to recognize animals with YOLO in a farming application

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 45

6.8.6 Disadvantages of YOLOv3 vs. Other Algorithms

The YOLOv3 AP does indicate a trade-off between speed and accuracy for using
YOLO when compared to RetinaNet since RetinaNet training time is greater than YOLOv3.
However, the accuracy of detecting objects with YOLOv3 can be made equal to the accuracy
when using RetinaNet by having a larger dataset, making it an ideal option for models that can
be trained with large datasets.

An example of this would be common detection models like traffic detection, where
plenty of data can be used to train the model since the number of images of different vehicles
is plentiful. On the other hand, YOLOv3 may not be ideal for using niche models where large
datasets can be hard to obtain.

6.9 Previous methods

 6.9.1 Classical face recognition algorithms

There has been a rapid development of the reliable face recognition algorithms in the last
decade. The traditional face recognition algorithms can be categorised into two categories:
holistic features and local feature approaches. The holistic group can be additionally divided
into linear and nonlinear projection methods.

Many applications have shown good results of the linear projection appearance‐based
methods such as principal component analysis (PCA), independent component analysis (ICA),
linear discriminate analysis (LDA), 2DPCA and linear regression classifier (LRC).

However, due to large variations in illumination conditions, facial expression and other
factors, these methods may fail to adequately represent the faces. The main reason is that the
face patterns lie on a complex nonlinear and non‐convex manifold in the high‐dimensional
space.

In order to deal with such cases, nonlinear extensions have been proposed like kernel PCA
(KPCA), kernel LDA (KLDA) or locally linear embedding (LLE). The most nonlinear
methods using the kernel techniques, where the general idea consists of mapping the input
face images into a higher‐dimensional space in which the manifold of the faces is linear and
simplified. So the traditional linear methods can be applied.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 46

Although PCA, LDA and LRC are considered as linear subspace learning algorithms, it is
notable that PCA and LDA methods focus on the global structure of the Euclidean space,
whereas LRC approach focuses on local structure of the manifold.

These methods project face onto a linear subspace spanned by the eigen face images. The
distance from face space is the orthogonal distance to the plane, whereas the distance in face
space is the distance along the plane from the mean image. These both distances can be turned
into Mahalanobis distances and given probabilistic interpretations.
Following these, there have been developed: KPCA, kernel ICA, and generalised linear
discriminant analysis.

Despite strong theoretical foundation of kernel‐based methods, the practical application of

these methods in face recognition problems, however, does not produce a significant
improvement compared with linear methods.

Fig: Traditional Face Recognition Algorithm

6.9.2 Artificial Neural Networks in face recognition

In artificial neural networks are used to solve nonlinear problem. To recognise human
faces, a non‐convergent chaotic neural network is suggested.
A radial basis function neural network integrated with a non‐negative matrix factorisation
to recognise faces is presented. Moreover, for face and speech verifications, utilise a
momentum back propagation neural network. Non‐negative sparse coding method to learning
facial features using different distance metrics and normalised cross‐correlation for face
recognition is applied.
A posterior union decision‐based artificial neural network approach is proposed. It has
elements of both neural networks and statistical approaches and replenishes methods for
recognising face images with partial distortion and occlusion.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 47

Unfortunately, this approach, like other statistical‐based methods, is inaccurate to model

classes given only a single or a small number of training samples.

6.9.3 Gabor wavelet‐based solutions

Gabor wavelets have been widely used for face representation by face recognition
researchers, and Gabor features are recognised as better representation for face recognition in
terms of (rank‐1) recognition rate. Moreover, it is demonstrated to be discriminative and
robust to illumination and expression variations. When only one sample image per enrolled
subject is available, propose adaptively weighted sub‐Gabor array for face representation and
recognition.
Moreover, two kinds of strategies to capture Gabor texture information: Gabor
magnitude‐based texture representation (GMTR) and Gabor phase‐based texture
representation (GPTR), are proposed.

Gamma density to model the Gabor magnitude distribution characterises GMTR

approach. The GPTR is characterised by the generalised Gaussian density for modelling the
Gabor phase distribution. It allows the estimated model parameters to be served as texture
representation of the face.

The Gabor wavelet applied at fixed positions, in correspondence of the nodes of a

square‐meshed grid superimposed to the face image, is presented. Each subpattern of the
partitioned face image is defined as the extracted Gabor features that belong to the same row
of the square‐meshed grid which are then projected to lower dimension space by Karhunen–
Loeve transform. The obtained features of each subpattern, which are weighted using genetic
algorithm (GA), are used to train a Parzen Window Classifier. Finally, matching process is
done by combining the classifiers using a weighted sum rule.

6.9.4 3D‐based face recognition

As 3D capturing process is becoming cheaper and faster, it is commonly thought that
the use of 3D sensing has the potential for greater recognition accuracy than 2D. The
advantage behind using 3D data is that depth information does not depend on pose and
illumination, and therefore, the representation of the object does not change with these
parameters, making the whole system more robust. 3D‐based techniques can achieve better
robustness to pose variation problem than 2D‐based ones. A comprehensive survey of the 3D
face recognition approaches is presented.
A method for face recognition across variations in pose, which combines deformable
3D models with a computer graphics simulation of projection and illumination, can be found.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 48

In this method, faces are represented by model parameters for 3D shape and texture. Their 3D
morphable models are combined with spherical harmonics illumination representation to
recognise faces under arbitrary unknown lighting.
Using facial symmetry to handle pose variation in 3D face recognition is presented,
where an automatic landmark detector is used. It helps to estimate pose and detects occluded
areas for each facial scan. Subsequently, an annotated face model is registered and fitted to the
scan. During fitting, facial symmetry is used to overcome the challenges of missing data.
There is a generic 3D elastic model for pose invariant face recognition proposed. It is
constructed for each subject in the database using only a single 2D image by applying the 3D
generic elastic model (3DGEM) approach. Each 3D model is subsequently rendered at
different poses within a limited search space about the estimated pose, and the resulting
images are matched against the test query. Finally, the distances between the synthesised
images and test query are computed by using a simple normalised correlation matcher to show
the effectiveness of the pose synthesis method to real‐world data.

6.9.5 Video‐based face recognition

The analysis of video streams of face images has received increasing attention in
biometrics. An immediate advantage in using video information is the possibility of
employing redundancy present in the video sequence to improve still image systems.
Although significant amount of research has been done in matching still face images, the use
of videos for face recognition is relatively less explored. The first stage of video‐based face
recognition (VFR) is to perform re‐identification, where a collection of videos is cross ‐
matched to locate all occurrences of the person of interest.
Generally, VFR approaches can be classified into two categories based on how they
leverage the multitude of information available in a video sequence: (i) sequence based and
(ii) set based, where at a high level, what most distinguishes these two approaches is whether
or not they utilise temporal information.
The formulation of a probabilistic appearance‐based face recognition approach is
extended. Originally, it was defined to do recognition from a single still image as previously
explained, to work with multiple images and video sequences. There is the constrained
subspace spanned from face images of a clip into a convex hull and then calculate the nearest
distance of two convex hulls as the between‐set similarity. Thus, each test and training
example is a set of images of a subject's face, not just a single image, so recognition decisions
need to be based on comparisons of image sets.
6.10 ADVANTAGES AND DISADVANTAGES

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 49

No
Method  Advantages  Disadvantages 
. 
These methods may fail to
adequately represent faces when
large variations in illumination facial
Focuses on local structure
expressions and other factors occur. Regarding
of the manifold.
applying kernel‐based nonlinear methods do
These methods project
not produce a
face onto linear subspace
significant improvement comparing to linear
spanned by the eigenface
methods. LLE, LLP and LBP brought simple
Classical face images. The distance from
and effective
1. recognition face space is
way to describe neighbouring changes in face
algorithms orthogonal to the plane
description. Subspace approaches
of mean image, so may be
were applied in DCV‐ and SVM‐based
easily turned to
methods. Preserving the local structure
Mahalanobis distances
between samples is the domain of NPP and
with probabilistic
ONPP methods.
interpretation
The problem is that it is still
unclear how to select the neighbourhood size
or assign optimal values for them
2. Artificial neural Radial basis function The main disadvantage
networks artificial of this approach is
neural network is naturally requirement of greater number of
integrated training samples (instead one or limited
with non‐negative number). It is inaccurate in the
matrix factorisation. same way like other statistically based methods
Also other
approaches for process
simplification regarding to
ANNs native linearisation
feature and computation
speed up.
Ideal solution, especially
for recognising face
images with partial

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 50

No
Method  Advantages  Disadvantages 
. 
distortion and occlusion
The drawback of the
The Gabor wavelets exhibit Gabor‐based methods is
desirable characteristics of significantly high dimensionality of
capturing the Gabor feature space since
salient visual properties like face image is convolved with a
spatial localisation bank of Gabor filters.
3. Gabor wavelets
orientation Approach is computationally
selectivity and spatial intensive and impractical for real‐time
frequency. Different applications.
biometrics applications Additionally, simplified
favour this approach Gabor features are sensitive
to lightning variations
The main idea behind
developing
image descriptors is to
learn
the most discriminant local
features that minimise
difference between
images of the same
individual Approach is computationally intensive during
Face descriptor‐ and maximise that between descriptor extraction stage, but encouraging
4.
based methods images from the other simplicity and performance in reference to
people. online applications
These methods are
discriminative and robust
to illumination and
expression
changes. They offer
compact,
easy to extract and highly
discriminative descriptor
5. 3D‐based Extend traditional Require all the elements of the 3D face

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 51

No
Method  Advantages  Disadvantages 
. 
2D capturing process
and has greater
potential for accuracy.
recognition system to be well calibrated and
The depth information does
face synchronised to existing 3D data.
not
recognition Computationally expensive and not suitable for
depend on the pose
practical applications
and illumination
making solution more
robust
The main advantage
of the approach is
possibility of employing Relatively poorly investigated.
Video‐based
6. redundancy Multiply problems with measuring similarity
recognition
present in video of two (or more) images
to improve still image
systems

6.11 Face recognition applications

Many published works mention numerous applications in which face recognition
technology is already utilised including entry to secured high‐risk spaces such as border
crossings as well as access to restricted resources. On the other hand, there are other
application areas in which face recognition has not yet been used. The potential application
areas of face recognition technology can be outlined as follows:
 Automated surveillance, where the objective is to recognise and track people.

 Monitoring closed circuit television (CCTV), the facial recognition capability can be
embedded into existing CCTV networks, to look for lost children or other missing
persons or tracking known or suspected criminals.
 Image database investigations, searching image databases of licensed drivers, benefit
recipients and finding people in large news photograph and video collections, as well
as searching in the Facebook social networking web site.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 52

 Multimedia environments with adaptive human computer interfaces (part of ubiquitous

or context aware systems, behaviour monitoring at childcare or centres for old people,
recognising customers and assessing their needs).
 Airplane‐boarding gate, the face recognition may be used in places of random checks
merely to screen passengers for further investigation. Similarly, in casinos, where
strategic design of betting floors that incorporates cameras at face height with good
lighting could be used not only to scan faces for identification purposes, but possibly
to afford the capture of images to build a comprehensive gallery for future watch‐list,
identification and authentication tasks.
 Sketch‐based face reconstruction, where law enforcement agencies in the world rely
on practical methods to help crime witnesses reconstruct likenesses of faces. These
methods range from sketch artistry to proprietary computerised composite systems.
 Forensic applications, where a forensic artist is often used to work with the eyewitness
in order to draw a sketch that depicts the facial appearance of the culprit according to
his/her verbal description. This forensic sketch is used later for matching large facial
image databases to identify the criminals. Yet, there is no existing face recognition
system that can be used for identification or verification in crime investigation such as
comparison of images taken by CCTV with available database of mugshots. Thus,
utilising face recognition technology in the forensic applications is a must as
discussed.
 Face spoofing and anti‐spoofing, where a photograph or video of an authorised
person's face could be used to gain access to facilities or services. Hence, the spoofing
attack consists in the use of forged biometric traits to gain illegitimate access to
secured resources protected by a biometric authentication system. It is a direct attack
to the sensory input of a biometric system, and the attacker does not need previous
knowledge about the recognition algorithm. Research on face spoof detection has
recently attracted an increasing attention, introducing few number of face spoof
detection techniques. Thus, developing a mature anti‐spoofing algorithm is still in its
infancy and further research is needed for face spoofing applications.
 CHAPTER-7

SYSTEM ANALYSIS

Prediction of terrorism activities is an important area of concern for researchers. The

large number of events makes it difficult to predict terrorist group responsible for some
terrorist activity.
Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
 CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 53

The current research is focused on finding out the correlation between terrorism and its
causal factors. Existing efforts have not been good enough for prediction. Machine learning
approaches can ad in predicting the likelihood of a terrorist attack, given the required data.
The results of this work can help the security agencies and policy makers to eradicate
terrorism by taking relevant and effective measures.

Hence there is an approach to analyzing terrorism region and country with the machine
learning techniques and terrorism specific knowledge to fetch conclusions about terrorist
behavior patterns.

7.1 Functional Requirements

The particular necessities are user interfaces. The outside clients are the customers. Every
one of the customers can utilize this product for ordering and looking.

 Hardware Interfaces: The outside equipment interface utilized for ordering and looking is
PCs of the customers. The PC's might be portable PCs with remote LAN as the web
association gave will be remote.
 Software Interfaces: The working Frameworks can be any rendition of windows.
 Performance Prerequisites: The PC's utilized must be at least pentium 4 machine with the
goal that they can give ideal execution of the item.

7.2 Non-Functional Requirements

Non utilitarian necessities are the capacities offered by the framework. It incorporates time
imperative and requirement on the advancement procedure and models. The non useful
prerequisites are as per the following:

 Speed: The framework ought to prepare the given contribution to yield inside fitting time.
 Ease of utilization: The product tought to be easy to understand. At that point the clients
can utilize effortlessly, so it doesn't require much preparing time.
 Reliability: The rate of disappointments ought to be less then just the framework is more
solid.
 Portability: It thought to be anything but difficult to actualize in any framework

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 54

CHAPTER-8

TEST RESULTS

TEST 1: Child is safe.

 Ultrasonic sensor measures the distance between child and parent.

 Distance in cm is displayed on LCD screen and serial monitor.
 Distance is less than safety limit hence buzzer is turned off.
 Safety message is shown on the LCD display.
 GPS module sends the latitude and longitude positions to Arduino Uno.
 Location data is displayed on the LCD screen.

TEST 2: Child is not safe.

 Ultrasonic sensor measures the distance between child and parent.

 Distance in cm is displayed on LCD screen and serial monitor.
 Distance is more than safety limit hence buzzer is turned ON to alert the
parent.
 Message for child is apart from parent is shown LCD display.
 GPS module sends the latitude and longitude positions to Arduino Uno.
 Location data is displayed on the LCD screen.
 GPS location data is sent to parent’s smartphone with location link through
SMS message service from GSM module.
 Google maps application is used to track the child location.
 Camera takes the image of the person caring the child.
 This image is processed through python and sent to parent through email
service.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 55

CHAPTER-9

ADVANTAGES

• Know the current location. Our Kids GPS Tracker provides real-time location of

your children.

• Get travel details of kids at any time.

• Emergency call.

• Message Alerts.

• Track your child even in a crowd.

• Uses of GPS Trackers for your Kids.

• Image capture for recognizing the caretaker and surroundings.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 56

CHAPTER-10

CONCLUSION

The child safety wearable device is capable of acting as a smart device. It provides
parents with the real-time location and SOS light along with Distress alarm buzzer for their
child's is 2m or more than that distance far and the ability to locate their child or alert
bystanders in acting to rescue or comfort the child. The smart child safety wearable can be
enhanced much more in the future by using highly compact Arduino modules such as the
Arduino which can be sewed into fabrics. Also a more power efficient model will have to be
created which will be capable of holding the battery for a longer time.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 57

REFERERENCES

[1] Aditi Gupta, Vibhor Harit- Department of Computer Science & Engineering, North India
Institute of Technology, Najibabad, UP, India “Child Safety & Tracking Management System
by Using GPS, Geo-Fencing & Android Application: An Analysis” 18 August 2016.

[2] H. Moustafa, H. Kenn, K. Sayrafian, W. Scanlon and Y. Zhang, "Mobile wearable

communications [Guest Editorial]," in IEEE Wireless Communications, vol. 22, no. 1, pp. lO-
l1, February 2015.

[3] S. Nasrin and P.1. Radcliffe, "Novel protocol enables DIY home automation,"
Telecommunication Networks and Applications Conference (ATNAC), 2014 Australasian,
Southbank, VIC, 2014, pp. 212-216.

[4] F. A. Silva, "Industrial Wireless Sensor Networks: Applications, Protocols, and Standards
[Book News]," in IEEE Industrial Electronics Magazine, vol. 8, no. 4, pp. 67-68, Dec. 2014.

[5]Jun Zheng; Simplot-Ryl, D.; Bisdikian, c.; Mouftah, H.T., "The internet of things [Guest
Editorial]," in Communications Magazine, IEEE , vo1.49, no.ll, pp.30-31, November 2011
doi: 10.1109/MCOM.2011.6069706.

[6]K. Braam, Tsung-Ching Huang, Chin-Hui Chen, E. Montgomery, S. Vo and R. Beausoleil,

"Wristband Vital: A wearable multi-sensor microsystem for real-time assistance via low-
power Bluetooth link," Internet of Things (WF-IoT), 2015 IEEE 2nd World Forwn on, Milan,
2015, pp. 87-9l. doi: 10.l109/WF-IoT.2015.7389032.

[7] "Digital parenting: The best wearables and new smart baby monitors. The latest smart
baby monitors and connected tech for your peace of mind,'Tech. Rep., 2015. [Online].
Available: http://www.wareable.com/parenting/the-best -wearab lesbabies- smart-baby-
monitors.

. [8] "WiFi and WiMAX - breakthrough in wireless access technologies," Wireless, Mobile
and Multimedia Networks, 2008. lET International Conference on, Beijing, 2008, pp. 141-
145.
Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 58

[9] P. Bhagwat, "Bluetooth: technology for short-range wireless apps," in IEEE Internet
Computing, vol. 5, no. 3, pp. 96-103, May/Jun 200l.

[lO] Y. A. Badamasi, "The working principle of an Arduino," Electronics, Computer and

Computation (ICECCO), 2014 11th International Conference on, Abuja, 2014, pp. 1-4.

[11] N. N. Prince, "Design and implementation of Arduino based short message service
control system," Internet Technology and Secured Transactions (ICITST), 2013 8th
International Conference for, London, 2013, pp. 494-499.

[12] A. Anastasiou, C. Tsirmpas, A. Rompas, K. Giokas and D. Koutsouris, "3D printing:

Basic concepts mathematics and technologies," Bioinformatics and Bioengineering (BIBE),
2013 IEEE 13th International Conference on, Chania, 2013, pp. 1-4.

[13] B. Dorsemaine, 1. P. Gaulier, 1. P. Wary, N. Kheir and P. Urien, "Internet of Things: A

Definition and Taxonomy," Next Generation Mobile Applications, Services and
Technologies, 2015 9th International Conference on, Cambridge, 2015, pp. 72- 77.

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 59

PHOTOS

Compilation of program in arduino software

When the object is far (ie at distance 17)

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 60

When the object is near (ie at distance 2)

Running the program in python software

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 61

Output of the program in python software

Location link sent to the number stored through sms

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru
CHILD TRACKING DEVICE USING YOLOV3 TECHNOLOGY Page 62

Image sent through the mail

Dept. of Electrical and Electronics Engineering Sri Sairam College of Engineering, Bengaluru