Closed-Loop Systems of Line Follower Robot Using Camera Vision

Closed-Loop Systems of Line Follower Robot Using

Camera Vision
W. H. M. Saad1*, M. N. Habib1, Z. M. Noh1, Z. Manap2 and M. S. J. A. Razak3
1
Fakulti Kejuruteraan Elektronik dan Kejururteraan Komputer, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, MALAYSIA
2
Fakulti Teknologi Kejuruteraan Elektrik dan Elektronik, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, MALAYSIA
3
SolokFertigasi by Perwira MSJ Enterprise, 75460 Duyong, Melaka, Malaysia.

*[email protected]

Abstract—Computer vision of robotic such as precision agriculture, healthcare and

applications has given a lot of benefits to robotic construction [1]–[4]. A line follower robot is
advancement. This study aims to design a robot mostly a robot designed to follow a line or path
that can follow a dark line path using a camera already fixed by the user. This line or path may
vision. A vision-based is a new technology that be as modest as a physical black line on the
uses an image processing technique instead white surface [5], [6] or as complex path marking
of using a sensor. An effective closed-loop structures, for example, embedded lines, and it
control system is used to overcome the robot's also can be unseen like a magnetic field [7], [8].
problems, moving out from the track by using Various sensing schemes can detect these specific
a line searching mechanism. This project used markers or 'lines'. These schemes may vary from
NI-myRIO as an embedded platform for image simple, low-cost line sensing circuits or moderate
processing and robot motion. The NI-myRIO vision systems [9]. These schemes' choices would
is connected to LabVIEW on PC wirelessly for depend upon the sensing accuracy and flexibility
broader processing capability while providing required.
maximum mobility to the robot to move freely From the industrial point of view, the line
on the designated track. Aiming for the guided following robot has been implemented in semi to
line to always be at the center of the region fully autonomous because the robot can replace
viewing window, each tire's speed on the left a human's job in the industry and can do many
and right part of the mobile robot is controlled things faster than humans [10]. Robots do not
accordingly. The result shows that using a line need to be paid for their work; they can do tedious
searching algorithm with a camera vision line work that is boring to people, and they will not
follower robot improves an overall mobile stop, slow down, or fall asleep like a human.
robot performance to follow the line without Individual stationary sensors have limited ranges
failure with faster speed. and applications.
Robotic site security sentries can work long
Keywords—Robot vision; Line follower robot; hours at a consistently high level of precision and
Line searching
vigilance [11]. In this environment, these robots
function as materials carriers to deliver products
I. INTRODUCTION from one manufacturing point to another where

R
rail, conveyor and gantry solutions are impossible.
OBOTICS development and application
Apart from the line following capabilities, these
using camera vision has accomplished
robots should also be able to navigate junctions
its extreme achievement in the world of
and decide which junction to turn and which
manufacturing and other fields of application
junction to ignore. This situation would require the
robot to have 90-degree turn and junction counting
Article history: Manuscript received 17 January 2022; received in
revised form 7 Mac 2022; Accepted 9 Mac 2022. capabilities.

ISSN: 2590-3551 eISSN: 2600-8122 Vol. 6 No. 1 April 2022 15

 configuration for basic robotic functions such as robot mobility
and motion. The second part is the development of lines
International Journal of Human and Technology Interaction
following an algorithm, including line detection and decision
making.
The development of robot structure emphasises the robot's
Line following robots with pick and direction are toward the centre of the detected line.
stability in motion to avoid the robot from falling while moving
placement capabilities is commonly used in an The block diagram in LabVIEW was developed,
around. Following algorithm development, the camera's image
industrial unit [12]. These move on a specified path streamand a line following
is pre-processed on program
the line to was created whether
determine using the
to pick the components from specified locations theposition
current line searching
and robotmechanism
direction are programmed in NI- of the
toward the centre
and place them in desired locations. In precision myRIO.
detected line. The block diagram in LabVIEW was developed,
agriculture, such a robot with vision capability and a line following program was created using the line
might be able to do the fruit picking task or de- searching
A. mechanism
Development programmed
of Line Followerin NI-myRIO.
Robot Structure
weeding the plant's medium [13]. A line followingA. In this stage, the
Development line follower
of Line Followersystem for the robot
Robot Structure T
robot is a self-operating robot that detects and Inis this developed. Thisline
stage, the stage includes
follower a line for
system follower's
the robot is
follows a line drawn on the floor. A black line basic structure
developed. This stageand working
includes a lineprinciple
follower'susingbasicNI-
structure IM
indicates the path to be taken on a white surface. myRIO principle
and working as the main controller.
using NI-myRIO as the main controller.
The control system must detect the line and ensure
the robot stays on course while correcting the
wrong moves. In this study, the development of a IM
vision-based line follower robot using NI-myRIO
is demonstrated to replace the traditional method
of using passive infrared (PIR) sensor previously
done in [5], [14].
The line searching mechanism is chosen to
prevent the robot from moving out from the track.
Thus, the robot moves more accurately with good IM
velocity on its track. On the other hand, another
feature like an obstacle detection system has been
introduced to prevent users from hitting any
Fig 1:FigWorking
1: Workingprototype
prototype ofofthe
the
lineline follower
follower robot robot
objects facing the robot on its track. This navigation
system is also used to navigate the robot back to The line follower system in this mobile system is guided
The line follower system in this mobile
the track in case the robot gets out from the track.by a webcam to follow the track line. The webcam is placed at
system is guided by a webcam to follow the track
This manuscript consists of four major the centre top of the robot so thatatit the
cancentre
capture and
line. The webcam is placed top of detect
the the
sections, including this section. The following black line more precise and clearer. Fig 1 shows the prototype
robot so that it can capture and detect the black
section is the methodology. In that section, the of the line follower robot. IM
line more precise and clearer. Figure 1 shows the
development process of the line following robot isB. Development of Line Following Robot Algorithm
prototype of the line follower robot.
explained in detail. Section III focuses on the result The communication between NI-MyRIO embedded device
and discussion; the last section is the conclusion.and LabVIEW software on the PC is through a Wi-Fi
B. Development of Line Following Robot Algorithm
connection
The where the programming
communication between of the NI-MyRIO
mobile robot is
taking place on both the PC and the embedded device. A few
embedded device and LabVIEW software on
II. METHODOLOGY Virtual Instruments (VI)s blocks were used in the line detecting
systemthedevelopment,
PC is through sucha Wi-Fi
as the connection
'IMAQdx Open where the VI'
Camera IM
Robot development is divided into two-part; the
programming
'IMAQdx Configure of the mobile robot
Acquisition is taking
VI', 'IMAQdxplace Start
first part is the development of robot structure,
on bothVI',
Acquisition the and
PC and the embedded
'IMAQdx device. A VI'.
Stop Acquisition few Fig 2
wiring and I/O configuration for basic robotic
showsVirtual
the block diagram to
Instruments display
(VI)s blocks an were
imageusedcaptured
in thefrom a
functions such as robot mobility and motion. The webcamlineondetecting
a GUI display.
system development, such as the
second part is the development of lines following fro
'IMAQdx Open Camera VI' 'IMAQdx Configure
an algorithm, including line detection and decision im
Acquisition VI', 'IMAQdx Start Acquisition VI', filt
making.
and 'IMAQdx Stop Acquisition VI'. Figure 3 shows cap
The development of robot structure
the block diagram to display an image captured
emphasises the robot's stability in motion to avoid
from a webcam on a GUI display. 'IM
the robot from falling while moving around.
Following algorithm development, the camera's ISSN: 2590-
image stream is pre-processed on the line to
determine whether the current position and robot

16 ISSN: 2590-3551 eISSN: 2600-8122 Vol. 6 No. 1 April 2022

 action (IJHaTI),
raction
raction (IJHaTI), Vol.
(IJHaTI), Vol. 6,
Vol. 6, No.
6, No. 1,
No. 1, 111 April
1, April 2022
April 2022
2022

Closed-Loop Systems of Line Follower Robot Using Camera Vision

first part (IJHaTI), Vol. 6, No. 1, 1 April 2022
nteraction
first
first part
first part
part
and I/O
and
and I/O I/O
and
mobility I/O International
Particle Filter'Journal
are beingofimplemented
Human and into
Technology
the line Interac
mobility
mobility
mobility
of lines International
International
detecting block Journal
Journal of
of
circuit.of Human
Human
Table and
and
2 shows Technology
Technology Interac
Interac
he offirst
of
of lines
lines
lines
part International Journal Human and the function Interac
Technology
gdecision
and I/O
bot
decision
decision
decision
mobility andof filter
'IMAQ blocks.Particle Filter' are being implemented into th
entrobot's
of lines and
detecting
and 'IMAQ The
'IMAQ output
block
'IMAQ ParticleParticle
Particle imageFilter'
circuit. from
Filter' are
Table
are thebeing
2'IMAQdx
being shows implemented Extract
the
implemented function into
intoof th
th
eee robot's
nd robot's
decision
robot's andSingle Color Planes'
Filter'
is
are
connected
being
to the
implemented
'Image Src'
into th
moving detecting
blocks.
detecting block block circuit.
block circuit.circuit. Table Table 2
Table 22 shows shows
shows the the function
the function
function of of
of
ee'smoving
moving
moving
image detecting
s'sthe
a's image
image
robot's terminal
blocks.
blocks. The outputof 'IMAQdx image from Threshold'
the 'IMAQdx to replace Extract the Single
a's
ether imagethe blocks.
hile
ether
ether moving
the
the value
Planes' The
Theofisoutput
the
connected
output image
pixels
image to from
existing the
from the
in the
'Image
the 'IMAQdx
range
'IMAQdx Src' Extract
between
terminal
Extract Single
of
Single 'IM
ether
reera's
re of
the
of image
the
of the Fig 2: Block diagram to display the image captured Planes' by The output image from the 'IMAQdx Extract Single
tre
tre
veloped,of the
the Fig
Fig 2:
2: Block
Block diagram
diagram to
to display
display the
the image
image captured
captured lower value
Threshold'
Planes'
by
by is
is connected
to and upper
replace
connected to
the
to the
value
the 'Image
value. of
'Image TheSrc'
the upper
pixels
Src' terminal
terminal and of
existing of in 'IM
the
'IM
whether
veloped,
veloped,
the Fig 2: Block diagram thetowebcamdisplay the image capturedPlanes' by is connected to the 'Image Src' terminal of 'IM
veloped,
the
entrelineof the the
the webcam
webcam Threshold'
between
lower values
Threshold' to
lowerto replace
arevalue
replace fixedthe
theand value
in upper
this
value of the
value.
project,
of the pixelsfrom
pixels The existing
existingto in
0 upper inand the
the
the
the line
line Fig 2: Block diagram the
to webcam
display
Fig 3: Block diagram to display the image capturedthe image captured by Threshold' to replace the value of the pixels existing in the
developed,
the line the webcam between
values
93.
between are
'Image lower
lowerfixed
Out' value
in
terminal
value and
this
and ofupper
project,
'IMAQdx
upper value.
from
value. The
0
Threshold'
The to upper
93.
upper and
'Image
and
ng the line by the webcam between lower value and upper value. The upper and
values
terminal
values are
are fixed
of fixed
'IMAQdx
fixed in this project,
Threshold'
in 'image'
this project, from
arefrom 0 to
ofconnected 93. 'Imag
93.to'Image the 'i
Table 1: The function of basic Vision and Motion Block
Table
Table
Table 1:
1: The
The
1:The
The function
function
function of
of basic
basic
Vis
of basic Vision
Vision
Vision and
and
and Motion Motion
Motion Block
Block
Block
are connected
values are to the
in this terminal
project, from 00 to
'IMAQdx to 93. 'Imag
Table 1: function of
Table 1: The function of basicVis basic Vision
Vis and Motion Block Vis terminal
terminal
Overlay of of
of 'IMAQdx
'IMAQdx
'IMAQdx
Rectangle' Threshold'
Overlay
Threshold'
for Threshold'
specifies theare are
Rectangle'
are connected
connected
coordinates forof toto
to the the
the 'i''
specifie
robot
e
robot
robot is
is
is
Vision and Motion Block
Vis terminal 'IMAQdx connected
robot
structure is IMAQdx VI IMAQdx VI Vis Function
Function terminal
coordinates
terminal
the rectangle of
of of 'IMAQdx
the
'IMAQdx rectangle
to overlay Overlay Overlayto
Overlay Rectangle'
overlay
and haveRectangle'Rectangle' and
drawing mode have for
for specifi
drawing
specifi
he
structure robot is
structure IMAQdx
IMAQdx VI
IMAQdx VI
VI Function
Function
Function
Open the camera,such
terminal of 'IMAQdx for specifie
structure
ntroller.
ic structure IMAQdx VI Function
Open
Open thethe the camera,
camera, coordinates
coordinates
such as
as of
specifies
of
specifies the
the rectangle
whether
rectangle
whether to to
to
draw overlay
tooverlay
draw
overlay
only the and
only
and frame have
have drawing
theofdrawing
frame o
drawing
ntroller.
ontroller.
ntroller.
Open
OpenqueriesOpen
the camera, thethecamera,
camera coordinates of the rectangle to and have
controller. queries
queries
camera, the
the
queries camera
camera such
rectangle,
such as
as
the rectangle, specifies
fill
specifies in the whether
rectangle,
whether
fill inwhether
the rectangle, to
toor draw
highlight
draw
or highlightonly
only thethe
the frame
rectangle.
frame o
the frame
queries queries
for the
itscamera the
for camera
capabilities, such as specifies to draw only the
the
for its for
forcamera
its
its capabilities,
capabilities,capabilities,
a camera rectangle,
rectangle,
shows
rectangle, the
rectangle. fill fill
fill
Figure in
connection
in the
the rectangle,
between
rectangle,
4 shows or
or
the connection highlight
the
highlight 'IMAQbetween the
the rectangle.
Threshold
rectangle
for
itsload
load load
its
capabilities,
load
capabilities,
aaa camera
camera 'IMAQdx
in the rectangle, or highlight the rectangle.
IMAQdx OpenCameraCamera VI aload
camera
configuration camera shows
shows
fileshows the
theOverlayconnection
connection Rectangle' between
and block.
between the
the 'IMAQ 'IMAQ
'IMAQ Threshold
Threshold
IMAQdx
IMAQdx
IMAQdx
IMAQdxOpen
IMAQdx Open
OpenCamera
Open
Open Camera
Camera
Camera
VIVIVI VI
VI
load a camera
configuration
configuration
configuration file file file the 'IMAQ the connection
Threshold' between 'IMAQdx the Overlay Threshold
configuration
configuration
and create a file'IMAQdx
'IMAQdx Overlay
Overlay Rectangle'
Rectangle' block.
block.
andfile create
and
and
andandcreate
a
create
create
create aaaa 'IMAQdx
Rectangle'Overlay block. Rectangle' block.
unique unique
unique reference
unique
unique
reference
reference
reference
reference
Table 2: The function of filtering VIs
to the tounique
the
camera. reference
camera. Table 2: The function of filtering VIs
toto to thethe camera.
camera. Table
Table 2:function
2: The function
The function of
of filtering VIs
VIsfiltering VIs
low-a low- IMAQ Block VI Function
the camera. Table 2: The of filtering
Configure to the
Configure a camera.
Configure
Configure a a low-
low- IMAQ Block VI VI Function
Determine whether to
level Configure
acquisition
Configure
level a low-
acquisitiona low- IMAQ IMAQ
IMAQ Block Block
Block VI VI Function
Function
Function
level
level
previously level acquisition
acquisition
acquisition
opened
level
previously acquisitionopened Determine
replace
Determine
Determine the
whether whether
value
whether of to
the
to
IMAQdx Cinfigure Acquisition VI with previously
previously opened
previously
IMAQdx
previously opened
opened Determine whether to
IMAQdx Cinfigure Acquisition VI Open with
opened
with
with IMAQdx
IMAQdx
with replace
to pixels
replace
replace the the
existing
value
the value
value in of
the
of the
ran
the
of replace
the pixels the value of the
IMAQdx
IMAQdx Cinfigure
Cinfigure Acquisition
Acquisition VI
VI Camera
with IMAQdx VI.
IMAQdx Cinfigure Acquisition VI VI Open theIMAQdx
Camera
Open VI.
IMAQdx Cinfigure Acquisition existing
IMAQdx
Specify Open
Open Camera
Camera VI.
VI. pixels
between
pixels
in pixels
existing
the rangeexisting lower
existing
between in the ran
in
value
in the
the ra
ra
Open VI.
Specify
Camera
acquisition Camera
type the VI.
Specify
Specify the
the between
(default
lowerbetween
value (default lower
is 128)
lower value
and
value upp
Specify
with acquisition
Specify
the
acquisition
the the type IMAQ Threshold
type
between lower value
r robot acquisition
acquisition
acquisition
with
'Continuous' theand type IMAQ
type
Threshold (default
is 128)
value and upper
(default is
(default 128)
128)isand
isis 128) 225)upp
and upp
obot with
with the
the IMAQ Threshold (default
value (default is 225) and upp
obotobot numbertype with with
of
'Continuous' the
buffers'
the and IMAQ
IMAQ Threshold
Threshold
value
Specifies
value (default
(default the is
group
is 225)
225) to
m obot is guided 'Continuous'
'Continuous'
'Continuous'
parameters. andand and value (default is 225)
'Continuous'
number of buffers'and Specifies
Specifies
which
Specifies the group
onethe the
wants
the group
group to to to
add
to
sisguided placed at
Startsnumber an of buffers'
number
number
number of
of buffers'
of buffers'
buffers' to Specifies
which one wants group
sdss detect
guided
guided the parameters.
parameters. which
overlay one wants
information. to addIf
placed
placed e
guided at
prototype parameters.
acquisitionparameters.
parameters.
that which
to which
add one
one
the overlay wants
wants to
to add
add
placed at at Starts an overlay information. If
placed
etect the
etect
etect the
at
the IMAQdx Start Acquisition VI
wasStartspreviously
Starts
Starts an
Startswith
acquisition
an
an
anthatthat name
overlay
information.
overlay is specified,
information.
If no
information. the If
If
etect the configured name is specified, the
rototype acquisition
acquisition
acquisition that
that name
overlay
name is
is specified,
information
specified, the
the is
prototyperototype
hm
rototype IMAQdx Start Acquisition VI
thewas acquisition
was
IMAQdx
was
was
previously
previously
previously
previously
that nameinformation
overlay is specified, the
IMAQdx
IMAQdx Start
Start Acquisition
Acquisition VI VI VI Configure was
configured previously
withwith IMAQ Overlay Rectangle overlay
added
overlay
is added toto information
the default gro
information is
is
dded device IMAQdx Start
IMAQdx StartAcquisition
Acquisition VI configured
configured
configured VI. with with overlay theinformation is
h a Wi-Fi Acquisition
the configured
the IMAQdx
IMAQdx with IMAQ Overlay Rectangle added
It
default also
added group. to
to the
It default
specifies
the the
default gro
gro
m
mle robot is
m Stops the an IMAQdx
the
the IMAQdx
IMAQdx
IMAQ
IMAQ
Overlay
IMAQ Overlay Rectangle alsoadded
Rectangle
Overlay Rectangle specifies tothe the default gro
d device
Configure
Configure It also
coordinates
It also specifies
of the of the
also specifies
specifies the
therecta
ed vice.dd device
device
A few Configure
acquisitionConfigure
Acquisition
Configure
Acquisition VI.
VI. It
coordinates the
ane device coordinates
to overlay on of the recta
of theimage
the
Wi-Fi Acquisition
previously Acquisition started VI.
VI. coordinates
rectangle to overlay of the recta
aaa detecting
robot
Wi-Fi
Wi-Fi
Wi-Fi isVI' IMAQdx Stop Acquisition VI with Acquisition
Stops
Stopsthe an an
IMAQdx VI. coordinates
on totheoverlay
image. on the image
rectan
Camera
robot
robot is
is Stops
Stops
Stops
acquisition an
an
an Take
to
to overlay
overlayouts selected
on
on the
the imageplan
image
e.Qdx robot
A few is
Start Startacquisition
Acquisition
e. e. A A few
few acquisition
acquisition
previously
acquisition started TakeTake
Takeouts outs
colour models
selected
outs selected
selectedlike plan
RGB
plan
e.
detecting A
VI'. few
Fig 2 VI. previously
previously
previously
started
started
started Take outs
planes colour models selected plan
detecting
detecting
detecting
ured from VI' a IMAQdx
IMAQdx Stop Acquisition VI
IMAQdx Stop Acquisition VI with the IMAQdx
previously
with the IMAQdx started colour
(red,
colour models
green,
models blue),like
like RGB
HSLRGB
mera
meramera VI' VI' IMAQdx
IMAQdx Stop
Stop
Stop Acquisition
Acquisition
Acquisition on theVI VI
VI with
Startwith Block the
the IMAQdx
Acquisition IMAQdx likecolour
RGB (red,modelsgreen, like RGB
xmeraStart VI' The brief information Vision and Motions with
Start the VIs
IMAQdx
Acquisition (red,
(hue,
(red,
blue), HSLgreen, (hue, blue),
saturation,
green, blue), HSL
lightne
HSL
xxx Start Start from LabVIEW is shown in Table 1. Each of the
VI.Start
Start
Start Acquisition
Acquisition
VIsAcquisition
plays an (red, green, blue), HSL
'. FigStart2 important role in the line detecting system.VI.
VI.
VI. (hue,
saturation,
HSV
(hue, saturation,
lightness),
(hue,
saturation, saturation, lightne
lightne
I'.I'. Fig Fig 222 VI. Some image
IMAQ Extract Single Colour HSV (hue,(hue,saturation,
saturation, lightne
dI'.from Fig a filtering VIs are added to the system circuit to improve the HSV
value)
HSV (hue,
or
(hue, HSI saturation,
(hue,
saturation,
ddd from from aaa TheThe brief information on the Vision Block and VIs IMAQ Extract Single HSV
value) or HSI(hue, (hue, saturation,
from captured brief
imageinformation
quality. on the Vision and Motions Plane
IMAQ
IMAQ Extract Extract Single ColourColour
Single Colour
Plane value) or HSI (hue,
from
The
The brief
Motions
The brief
brief
LabVIEW
information
information
Block
informationVIs
istheshown fromon
on the
on
in
the Vision
Vision and
LabVIEW
the
Table Vision
1.ofEach
andisMotions
and of
Motions
shown
Motions
the VIsBlock
Block
Blockin VIs
plays VIsIMAQ
VIs
an Extract Single Colour saturation,
value) intensity)
saturation,
value) or HSI
or HSIintensity)
(hue, fro
(hue,
from
from
In this
LabVIEW
LabVIEW
study,
is
additional
is shown
shown'IMAQ in
filter
in Table
Table
VI
1.
'IMAQ
1. Each
Each of
Threshold',
the
ofColorthe VIs VIs plays
plays an Plane
Plane
an fromsaturation,
an an image intensity) fro
image
from
important LabVIEW
Table 1. role role
Each is
in shown
the linein Table
detecting 1. Each system. of the VIs
Some plays
image Plane
an saturation,
saturation, intensity) fro intensity) fro
inof thethe line
VIs detecting
plays an system.
important roleimage
'IMAQ Overlay Rectangle', Extract Single Plane'
important
important role in the line detecting system. Some Some image an image
important
filtering
in the line
filtering
filtering
ISSN: 2590-3551
VIsrole
are in
VIs detecting
VIs are
are added
the
added line
to
added system.
to
the
to the
detecting
system
Somecircuit
the system
system
system.
circuit
imageto
circuit
to
to
Some
improve
filtering
improve 2
improve the
image
the
the
Define
Define
an the Region
image
an image the Region
of of
Interest (RoI)within
filtering
captured VIs
image are added
quality. to the system circuit to improve the Define
Interest
Define the Region
(RoI)within
the Region of
ofwh
captured
VIsInare
captured
captured image
added
image
image quality.
to
quality.
quality. the system circuit
this study, the additional filter VI of 'IMAQ Threshold',
to improve the Define
which particle the filterRegion
is of
In
In this
this study,
study, the
the additional
additional filter
filter VI
VI of
of 'IMAQ
'IMAQ Threshold',
Threshold', Interest
particle
Interest
performed and (RoI)within
filter
(RoI)within
contain is perform wh
wh
captured
'IMAQ In this
Overlayimage
study, quality.
the
Rectangle',additional 'IMAQ filter VI ofSingle
Extract 'IMAQColor Threshold',
Plane' Interest
theparticle
coordinatefilter
(RoI)within
of the is perform
wh
'IMAQ
'IMAQ
'IMAQ

Overlay
Overlay
Overlay Rectangle',
Rectangle',
Rectangle', 'IMAQ
'IMAQ
'IMAQ
In this study, the additional filter VI of
Extract
Extract
Extract Single
Single
Single Color
Color
Color Plane'
Plane'
Plane' and
particle
particle contain
bounding rectangle to filter
filter theis
is coordin
perform
perform
ISSN: 2590-3551 2 and
of
andthe contain
bounding
contain the coordin
rectang
the coordin
coordin
ISSN:
ISSN: 'IMAQ
2590-3551
2590-3551 Threshold', 'IMAQ Overlay Rectangle', 222 and
indicate contain
the contour's the
ISSN: 2590-3551 of
of
relative the
indicate
the bounding
the
bounding
position. contour'srectang
rectang
'IMAQ Extract Single Color Plane' and 'IMAQ IMAQ Particle
IMAQ Particle FilterFilter of the bounding rectang
IMAQ Particle Filter indicate
relative
indicate the
position.
the contour's
contour's
IMAQ Particle
IMAQ Particle Filter Filter indicate the contour's
relative
relative position. position.
position.
relative
After the rectangular size has been selected, the
ISSN: 2590-3551 eISSN: 2600-8122 image Vol.of6 'IMAQdx
After No. rectangular
the 1 April 2022size
Overlay has
Rectangle' been selected,17 throug the
After
After the
the rectangular
rectangular size
size has been
has beenwill pass
selected,
selected, the o
the
image
image of
to
of 'IMAQdx
the
'IMAQdx 'IMAQ Overlay
Particle
Overlay Rectangle'
Filter'
Rectangle' block. will
will Whenpass
pass throu
anthrou im
image of 'IMAQdx Overlay Rectangle' will pass throug
 connected to the input session of 'IMAQdx Stop Acquisition' to
Specifies the group to stop an acquisition previously started with the 'IMAQdx Start
which one wants to add the Acquisition'. Next, the output session of 'IMAQdx Stop
International Journal of Humanoverlay
and Technology
information.Interaction
If no Acquisition' has been connected to the input session of
name is specified, the 'IMAQdx Close Camera' to stop an acquisition in progress,
overlay information is releases resources associated with an acquisition, and closes the
IMAQ Overlay Rectangle added to the default group.
After the rectangular size has been
It also specifies the
can useCamera
specified IMAQ Dispose
Session. foroutput
Then, the eachimagecallofto'IMAQdx
IMAQ
Get Image' has been connected to 'IMAQdx Dispose Vi' to
selected, the output image of 'IMAQdx
coordinates Overlay
of the rectangle Createanor
destroy justand
image oncefree for all images
the space created
it occupies with
in memory.
Rectangle' will pass through to overlay
theonimage
the image.to the IMAQIMAQ
Execute Create. Figure
Dispose only 6whenshows the connection
the image is no longer
Take outs selected planes needed in your application. You Acquisition',
can use IMAQ Dispose for
'IMAQ Particle Filter' block. Whenlike
colour models anRGB image between 'IMAQdx Stop 'IMAQdx
each call to IMAQ Create or just once for all images created
is processed through 'IMAQ (red, green,Particle
blue), HSL Filter', Close
with Camera'
IMAQ Create.and
Fig 'IMAQdx
5 shows the Dispose Vi'.between
connection Figure
(hue, saturation, lightness),
the output is masked, and only the Region of 7 showsStop
'IMAQdx the Acquisition',
Overall Block Diagram
'IMAQdx of the Line
Close Camera' and
HSV (hue, saturation, 'IMAQdx Dispose Vi'. Fig 6 shows the Overall Block Diagram
Interest (ROI)Single
IMAQ Extract from the image
Colour is HSI
value) or displayed
(hue, with Detection
of System.
the Line Detection System.
Plane saturation, intensity) fromout.
all the other particles of the image blackened
an image
The input image of the 'IMAQ Define Particle
the RegionAnalysis'
of
is connected to the output image
Interest of the which
(RoI)within 'IMAQ
particle
Particle Filter'; this block will filter isthe
return performed
number
and contain the coordinate
of particles detected in aof binarythe boundingimage andto a
rectangle
2D array
IMAQ of requested
Particle Filter indicate the contour's
measurements about the
relative position.
particle.
n (IJHaTI), Figure
Vol. 6, No. 5 shows
1, 1 April 2022the connection between
'IMAQ Particle
After the Filter'
rectangular and
size has been 'IMAQ Particle
selected, the output
imageFig
of 'IMAQdx
Analysis'.
ne particle. 4 shows Overlay Rectangle'between
the connection will pass'IMAQ
through Particle
the
image to the 'IMAQ Particle Filter' block. When an imageInternational
is Journal of Human and Technology Interaction (IJHa
ter Filter' and 'IMAQ Particle Analysis'.
processed through 'IMAQ Particle Filter', the output is masked,
and only the Region of Interest (ROI) from the image is
International
Fig Fig 4:Journal
5: Connection of between
Connection Human
between and
'IMAQ
'IMAQ Technology
Particle
Particle Interaction
Filter'Filter' and
and 'IMAQ (IJHa
'IMAQ Particle
Particle Analysis'
Analysis'
lor displayed with all the other particles of the image blackened m
dx out. The input image of the 'IMAQ Particle Analysis' is T
ge connected to the output image of the 'IMAQ Particle Filter'; this m
block will return the number of particles detected in a binary
L
T
wer
image and a 2D array of requested measurements about the L
ut'
ge' ISSN: 2590-3551 3 o
he op
Fig 5: Connection between 'IMAQdx Stop
de pb
Acquisition',5: 'IMAQdx Close Camera' and 'IMAQdx
he FigFig Connection
6: Connection between between
'IMAQdx 'IMAQdx
Stop Stop
Acquisition', bA
g3 Dispose
'IMAQdx Close Camera' VI'
and 'IMAQdx Dispose VI'
Acquisition', 'IMAQdx Close Camera' and 'IMAQdx 1
A
nd
Dispose VI' 1m
mw
wb
bc
cw
wT
Fig3:
Fig 4: Connection
Connection of of
thethe
'IMAQ Threshold'
'IMAQ and ' and
Threshold' s
IMAQdx Overlay
'IMAQdx Overlay Rectangle'
Rectangle' T
r
s
The The
outputoutput
session session
of 'IMAQdx Start Acquisition'
of 'IMAQdx Start is r
connected to the input
Acquisition' session of 'IMAQdx
is connected Stop Acquisition'
to the input session to
stop an acquisition previously started with the 'IMAQdx Start
of 'IMAQdx
Acquisition'. Next, Stop Acquisition'
the output session ofto 'IMAQdx
stop anStop
acquisition
Acquisition' haspreviously startedtowith
been connected the the 'IMAQdx
input session of
Start Acquisition'.
'IMAQdx Close Camera'Next, to stopthean output
acquisitionsession of Fig
in progress, 6:Overall
Fig 7: Overall
Block Block Diagram
Diagram of the LineofDetection
the Line Detection
System s
releases resources
'IMAQdx associated
Stop with anhas
Acquisition' acquisition, and closes the
been connected System li
s
specified Camera Session. Then, the output image of 'IMAQdxFig 6: Overall Block Diagram of the Line Detection m
to the input session of 'IMAQdx Close Camera' A robot movement system has been li
Get Image' has been connected to 'IMAQdx Dispose Vi' to System W
e to stop
destroy an imagean acquisition in progress,
and free the space it occupiesreleases
A robot movement system has been developed to navigate
in memory.developed to navigate the robot whenever the m
the robot whenever the linedetects
followera robot detects a line on its p
resources
Execute IMAQassociated
Dispose only withwhenan the
acquisition,
image is noand A line
longer robotfollower
movement robotsystem has been line on its path.
developed to navigate W
path. This movement system has been created using LabVIEW o
needed in your
closes application.Camera
the specified You canSession.
use IMAQ Then,Dispose forThis
the robot
the movement
whenever the system has been
line follower robotcreated
detectsusing
a line on its p
eachoutput
call to image
IMAQ of Create or just once for all
'IMAQdx Get Image' has been images software.
created Users can also manually control the robot's direction, o
path. This movement
LabVIEW systemUsers
software. has been cancreated using LabVIEW
also manually c
with IMAQ Create. Fig 5 shows the connection between
connected to 'IMAQdx Dispose Vi' to destroysoftware.
such as 'Left',
an andcontrol
'Right', 'Forward' and 'Backward'. Fig 7 shows the
'IMAQdx Stop Acquisition', 'IMAQdx Close Camera' Usersthecan robot's
also manually control
direction, the robot's
such direction,
as 'Left', c
manual navigator of the line follower robot. c
imageDispose
'IMAQdx and free the
Vi'. Figspace it occupies
6 shows the OverallinBlock
memory.such as'Right',
Diagram 'Left', 'Right', 'Forward'
'Forward' and and 'Backward'.Figure
'Backward'. Fig 7 shows
8 the m
manualshows navigator of the line followerofrobot. c
Execute
of the IMAQ System.
Line Detection Dispose only when the image the manual navigator the line follower r
m
is no longer needed in your application. You robot. s
r
tr
s
it
tr
18 ISSN: 2590-3551 eISSN: 2600-8122 Vol. 6 No. 1 April 2022
it
o
the robot whenever the line follower robot detects a line on its programmed to have one wheel slowed down to half the
path. This movement system has been created using LabVIEW original speed to ensure smooth turning.
software. Users can also manually control the robot's direction, However, during the preparation of the line path, one
Closed-Loop Systems of Line Follower Robot Using Camera Vision
such as 'Left', 'Right', 'Forward' and 'Backward'. Fig 7 shows the corner of the loop path is purposely designed to have a sharper
manual navigator of the line follower robot. corner to observe the robot's reaction. As a result, the camera
missed detecting the line at the designed sharp corner. The
robot's moving speed could not match the supposed turning
III. RESULT AND ANALYSIS
speed. Before the robot can finish a turn, it is already off-
A. and
tracked Line Following
leads Performance
to zero output coordinate of line detecting as
it detects no line. testing result of the prototype robot
The overall
B.was Linesatisfactory,
Searching Mechanism especially when it was being
The line searching mechanism is a method used to prevent
tested in a straight line. In the LabVIEW block
the line follower robot from moving out from the line. Using
Fig
Fig8:7:The
TheManual
ManualNavigator
NavigatorofofLine
LineFollower
FollowerRobot
Robot thediagram,
camera for thethe wheel
detect line ismode of the dc
a huge difference motor
from is set
using an
infrared
to the sensor because the
maximum camerawhich
speed, providesisa 0.5
wideduty
angle cycle.
of
Table 3: Pin Connection between NI-myRIO and L298N vision, so it will easily move out from its path. The line
Table 3: Pin Connection between NI-myRIO and
Driver
L298N Driver
Whenmechanism
searching the prototype
will help the robot sensesrobot
line follower theto find
corner
NI-myRIO L298N Driver its path,
track bythefinding the nearest
robot line available. This
is programmed tomethod
haveisone
Analogue A1NI-myRIO PWM L298N Driver
0 (pin 27) applied to the line following robot because it provides more
Pins
wheel slowed
advantages downmethod,
than the previous to halfsuch
theasoriginal
overcoming speed
the to
Analogue A2 A1 PWM PWM1 (pin
0 29)
(pin 27)
ensure
problem smooth
faced turning.
by all line following robots, which quickly step
Digital Pins
Pins B1 A2 DIO PWM
6 (pin 123)(pin 29) out from their track [15], [16].
B2 DIO 7 (pin 25)
WhenHowever, during the preparation of
the speed of the DC motor increases, it will affect
Digital B1 DIO 6 (pin 23) thethe line path,
probability of a robotone cornera lap.
completing of Inthe loop
this line path is
follower
Pins robot project, the speed of the dctomotor is controlled using pulse
B2 DIO 7 (pin
L298N H bridge is used as a driver to control the speed 25) of purposely designed have a sharper corner
the DC motor. Four pins have been used to connect NI-myRIO width modulation (PWM) by changing the duty cycle. An
and the L298N driver. Two analogue pins have been used to
to observe
analysis the by
has been done robot's
comparingreaction. Asmobile
the time the a result,
robot the
controlL298N H bridge
the Right Motor directionisandused
the LeftasMotor
a driver
direction.to took to complete
camera three laps
missed with different
detecting speeds
the line atofthe
DCdesigned
motor.
control
Analog pinthehelpspeed of athe
to generate pulseDC motor.
width modulationFour (PWM)pins
sharp corner. The robot's moving speed could
signal to an external peripheral through the PWM channels and
have been used to connect NI myRIO and the
not match the supposed turning speed. Before
change the duty cycle. Then, two digital pins are used to do the
L298N driver. Two analogue pins have been the robot can finish a turn, it is already off-
used to control the Right Motor direction and
ISSN: tracked and leads to zero output coordinate
2590-3551 4 of
the Left Motor direction. Analog pin help to line detecting as it detects no line.
generate a pulse width modulation (PWM)
signal to an external peripheral through the B. Line Searching Mechanism
PWM channels and change the duty cycle. The line searching mechanism is a method used
Then, two digital pins are used to do the manual to prevent the line follower robot from moving
navigation, such as turning on or off the DC out from the line. Using the camera for the
motor. Table 3 shows the connections between detect line is a huge difference from using an
the NI-myRIO and the L298N driver. infrared sensor because the camera provides a
The block diagram has been designed wide angle of vision, so it will easily move out
based on the centre of mass x. The camera will from its path. The line searching mechanism will
detect the line based on the range provided. The help the line follower robot to find its track by
centre of mass x is the centre of the rectangle finding the nearest line available. This method
box; this rectangle box will show the detected is applied to the line following robot because
line coordinate. A program has been designed it provides more advantages than the previous
in the centre of mass x in between 170 to 200, method, such as overcoming the problem faced
left and right dc motors will rotate clockwise by all line following robots, which quickly step
so the mobile robot will be moved forward. out from their track [15], [16].
The suitable DC motor will rotate clockwise When the speed of the DC motor increases,
whenever the centre of mass x coordinates it will affect the probability of a robot completing
between 201 to 500, so the mobile robot will turn a lap. In this line follower robot project, the
left when the centre mass coordinate between 1 speed of the dc motor is controlled using pulse
to 69, and the left dc motor will rotate clockwise width modulation (PWM) by changing the duty
so that the mobile robot will turn right. Thus, if cycle. An analysis has been done by comparing
the centre of mass is 0, the dc motor will stop. the time the mobile robot took to complete three
Figure 10 shows the overall front panel of the laps with different speeds of DC motor.
vision line following the robot.

ISSN: 2590-3551 eISSN: 2600-8122 Vol. 6 No. 1 April 2022 19

 International Journal of Human and Technology Interaction
International Journal of Human and Technology Interaction (IJHaTI), Vol. 6, No. 1, 1 April 2022

Fig10:
Fig 8:Overall
Overallfront
front panel
panel of vision
of vision lineline following
following robot
robot

International Journal of Human and Technology Interaction (IJHaTI), Vol. 6, No. 1, 01 April 2022

normal method. Thus, and

PJP/2020/FKEKK/PP/S01788 weproviding
could say thatof the
the use the
40
35
existing
line facilities tomechanism
searching complete this project.
works well in the line
An average of time taken for

30 following robot system.

25
REFERENCES
comple 3 laps

20
15 [1] K. H. Petersen, N. Napp, R. Stuart-Smith, D. Rus, and M. Kovac, "A review
of collective robotic construction," Science Robotics, vol. 4, no. 28, 2019.
10
IV. CONCLUSION
[2] N. S. Naik, V. V. Shete, and S. R. Danve, "Precision agriculture robot for
5 seeding function," in 2016 International Conference on Inventive
0 At Computation
the endTechnologies
of this(ICICT), project,
Aug. 2016,a vol.
fully
2, pp.functional
1–3.
[3] R. R Shamshiri et al., "Research and development in agricultural robotics:
0.15 0.2 0.35 0.45 0.5 prototype of a line follower robot using a camera
A perspective of digital farming," International Journal of Agricultural and
Speed of Normal
dc motor wasBiological
successfully
Engineering,designed and
vol. 11, no. 4, pp. 1–14,constructed.
2018. Line
[4] S. Shamsuddin, K. Argenan, S. R. Kamat, and M. H. Nordin, "Development
Fig follower robots
of a simple mobileare robotessential and vital,
for human-robot especially
interaction in
in health care
Fig 9:
12:The
Thecomparison
comparison between
between normal
normal and
and line searching
line searching environment," Journal of Advanced Manufacturing Technology (JAMT),
mechanism method
mechanism method industries, because they
vol. 12, no. 1(4), pp. 159–172, 2018. can help the worker carry
[5] W. H. M. Saad, S. A. A. Karim, N. Azhar, Z. Manap, Y. Y. Soon, and M.
the heavy load safely and more systematically.
M. Ibrahim, "Line Follower Mobile Robot for Surveillance Camera
By comparing the normal (switching method) and line
By mechanism,
searching comparing we couldthesay that
normal (switching
the performance line Besides
Monitoring that, theJournal
System," line offollower robot Electronic
Telecommunication, also can and
Computer Engineering (JTEC), vol. 10, no. 2–7, pp. 1-5–5, Jul. 2018.
searching mechanism is higher than the
method) and line searching mechanism, we could normal method. Fig 9 reduce
[6] F. the
Kaiser, number
S. Islam, W. of
Imran,hazards
K. H. Khan,in industries
and K. M. A. caused
Islam, "Line
shows the comparison between the normal and line searching follower robot: Fabrication and accuracy measurement by data acquisition,"
say that the
mechanism performance
method. Based on line
Fig 9,searching
the averagemechanism
time taken to
while carrying
in 2014 heavy
International loads on
Conference among workers.
Electrical EngineeringLineand
Information & Communication Technology, 2014, pp. 1–6.
three laps when the DC motor speed increased for the following robots is one
is higher than the normal method. Figure 12 shows [7] P. Ong, E. S. Low, and C. Y. Low, "Mobile Robot Path Planning
complete of the methods that provide
using Q-
line searching mechanism is more stable than the normal line
the comparison between the normal and line a safe environment.
Learning with Guided Distance and Moving Target Concept," International
following robot. This setup proved that the line searching Journal of Integrated Engineering, vol. 13, no. 2, Art. no. 2, 2021.
searching
mechanism could mechanism
prevent themethod.
line followerBased on Figure
robot from moving
[8] R. By
H. using
Ahrens, "Magnetic a lineline searching
follower for an xy motormechanism,
control," Apr. 07,
1970
out from the track. The line searching mechanism also produces an effectual output has
12, the average time taken to complete three laps [9] W. E. Elhady, H. A. Elnemr, and G. Selim, "Implementation and evaluation been obtained. From
more efficient motion and moves faster than the normal method. of image processing techniques on a vision navigation line follower robot,"
when
Thus, we the DCsay
could motor
that thespeed increased
line searching for theworks
mechanism line the analysis we identified, the line
Journal of Computer Science, vol. 10, no. 6, pp. 1036–1044, 2014. searching
searching
well in the linemechanism more stable than the mechanism
following robotissystem. [10] S. Robla-Gómez, is the
V. M.perfect
Becerra, J. method for a line follower
R. Llata, E. Gonzalez-Sarabia,
Ferrero, and J. Perez-Oria, "Working together: A review on safe human-
C. Torre-

normal line following robot. This setup proved robotrobot to overcome

collaboration in a mobile
industrial robot
environments," from
IEEE stepping
Access, vol. 5, pp.
IV. CONCLUSION 26754–26773, 2017.
that the line searching mechanism could prevent out from its track.
[11] Y. S. Yew, "Line follower for surveillance camera monitoring system,"
At the end of this project, a fully functional prototype of ISSN: a line 2590-3551 5
the linerobot
follower follower
using robot
a camera fromwas moving
successfullyout from the
designed and Journal This
10, no.
of Telecommunication, Electronic and Computer Engineering, vol.
2–7,project concept can be further integrated
pp. 1–5, 2017.
constructed. Line follower robots are
track. The line searching mechanism also produces essential and vital, into
[12] M.aA.stereo-vision
Rusafi, M. M. Billal, system in the
F. Yasmin, future
R. Khan, because
S. Ferdousi,
Iqbal, "A conceptual design of line follower pick-and-place robot for
the
and S. S.
especially in industries, because they can help the worker carry
more
the heavyefficient motion
load safely and moreand moves faster
systematically. Besidesthan
that,the
the
stereo camera
industrial can
purpose," in help track
Proceedings ofor detect
International lines better
Conference on
Mechanical, Industrial and Materials Engineering, 2017, pp. 1–6.
line follower robot also can reduce the number of hazards in [13] D. Robertson, M. Cook, E. Herbert, and F. Tully, "Robotic fruit picking
industries caused while carrying heavy loads among workers. system," Sep. 22, 2020
Line following robots is one of the methods that provide a safe [14] Z. Manap, W. H. M. Saad, H. R. Ramli, R. Marimuthu, and S. A. A. Karim,
20 ISSN: 2590-3551 eISSN: 2600-8122 Vol. 6 No. 1 April 2022
"Development of line follower robot with camera surveillance system,"
environment. IJRTE, vol. 8, no. 4, pp. 2192–2197, Nov. 2019.
By using a line searching mechanism, an effectual output [15] A. Attar, "Line follower and obstacle avoidance bot using Arduino,"
 Closed-Loop Systems of Line Follower Robot Using Camera Vision

by providing a wider view. This configuration [6] F. Kaiser, S. Islam, W. Imran, K. H. Khan, and
can prevent the mobile from moving out from K. M. A. Islam, "Line follower robot: Fabrication
the track because it cannot detect the line. Besides and accuracy measurement by data acquisition,"
that, using the stereo-vision system can precisely in 2014 International Conference on Electrical
detect lines, so the mobile robot can follow the line Engineering and Information & Communication
smoothly and accurately. Technology, 2014, pp. 1–6.
[7] P. Ong, E. S. Low, and C. Y. Low, "Mobile
Robot Path Planning using Q-Learning with
ACKNOWLEDGEMENT Guided Distance and Moving Target Concept,"
Authors would like to thank the Machine Learning International Journal of Integrated Engineering,
& Signal Processing (MLSP) research group under vol. 13, no. 2, Art. no. 2, 2021.
Fakulti Kejuruteraan Elektronik dan Kejuruteraan [8] R. H. Ahrens, "Magnetic line follower for an xy
Komputer (FKEKK), of Universiti Teknikal motor control," Apr. 07, 1970
Malaysia Melaka (UTeM) for sponsoring this
[9] W. E. Elhady, H. A. Elnemr, and G. Selim,
work under project PJP/2020/FKEKK/PP/S01788
"Implementation and evaluation of image
and providing the use of the existing facilities to
processing techniques on a vision navigation line
complete this project.
follower robot," Journal of Computer Science,
vol. 10, no. 6, pp. 1036–1044, 2014.
[10] S. Robla-Gómez, V. M. Becerra, J. R. Llata, E.
REFERENCES
Gonzalez-Sarabia, C. Torre-Ferrero, and J.
[1] K. H. Petersen, N. Napp, R. Stuart-Smith, D. Rus,
Perez-Oria, "Working together: A review on
and M. Kovac, "A review of collective robotic
safe human-robot collaboration in industrial
construction," Science Robotics, vol. 4, no. 28,
environments," IEEE Access, vol. 5, pp. 26754–
2019.
26773, 2017.
[2] N. S. Naik, V. V. Shete, and S. R. Danve, "Precision
[11] Y. S. Yew, "Line follower for surveillance
agriculture robot for seeding function," in
camera monitoring system," Journal of
2016 International Conference on Inventive
Telecommunication, Electronic and Computer
Computation Technologies (ICICT), Aug. 2016,
Engineering, vol. 10, no. 2–7, pp. 1–5, 2017.
vol. 2, pp. 1–3.
[12] M. A. Rusafi, M. M. Billal, F. Yasmin, R. Khan, S.
[3] R. R Shamshiri et al., "Research and development
Ferdousi, and S. S. Iqbal, "A conceptual design of
in agricultural robotics: A perspective of digital
line follower pick-and-place robot for industrial
farming," International Journal of Agricultural
purpose," in Proceedings of International
and Biological Engineering, vol. 11, no. 4, pp.
Conference on Mechanical, Industrial and
1–14, 2018.
Materials Engineering, 2017, pp. 1–6.
[4] S. Shamsuddin, K. Argenan, S. R. Kamat,
[13] D. Robertson, M. Cook, E. Herbert, and F. Tully,
and M. H. Nordin, "Development of a simple
"Robotic fruit picking system," Sep. 22, 2020
mobile robot for human-robot interaction in
health care environment," Journal of Advanced [14] Z. Manap, W. H. M. Saad, H. R. Ramli, R.
Manufacturing Technology (JAMT), vol. 12, no. Marimuthu, and S. A. A. Karim, "Development
1(4), pp. 159–172, 2018. of line follower robot with camera surveillance
system," IJRTE, vol. 8, no. 4, pp. 2192–2197, Nov.
[5] W. H. M. Saad, S. A. A. Karim, N. Azhar,
2019.
Z. Manap, Y. Y. Soon, and M. M. Ibrahim,
"Line Follower Mobile Robot for Surveillance [15] A. Attar, "Line follower and obstacle avoidance
Camera Monitoring System," Journal of bot using Arduino," International Journal of
Telecommunication, Electronic and Computer Advanced Computational Engineering and
Engineering (JTEC), vol. 10, no. 2–7, pp. 1-5–5, Networking, vol. 5, no. 4, pp. 18–21, 2017.
Jul. 2018.

ISSN: 2590-3551 eISSN: 2600-8122 Vol. 6 No. 1 April 2022 21

International Journal of Human and Technology Interaction

[16] C.-F. Hsu, C.-T. Su, W.-F. Kao, and B.-K. Lee,
"Vision-based line-following control of a two-
wheel self-balancing robot," in 2018 International
Conference on Machine Learning and
Cybernetics (ICMLC), 2018, vol. 1, pp. 319–324.

22 ISSN: 2590-3551 eISSN: 2600-8122 Vol. 6 No. 1 April 2022