IAES
Inter
national
J
our
nal
of
Articial
Intelligence
(IJ-AI)
V
ol.
15,
No.
4,
August
2026,
pp.
3614
∼
3624
ISSN:
2252-8938,
DOI:
10.11591/ijai.v15.i4.pp3614-3624
❒
3614
Smart
parking
management
system:
a
seamless
parking
solution
using
Y
OLO
and
QR
code
payment
technology
Sumit
K
umar
1
,
Ruchi
Rani
2
,
Sanjee
v
K
umar
Pippal
3
1
Department
of
Articial
Intelligence
and
Machine
Learning,
Symbiosis
Institute
of
T
echnology
Pune,
Symbiosis
International
(Deemed
Uni
v
ersity),
Pune,
India
2
Department
of
Computer
Engineering
and
T
echnology
,
School
of
Computer
Science
and
Engineering,
Dr
.V
ishw
anath
Karad
MIT
W
orld
Peace
Uni
v
ersity
,
Pune,
India
3
School
of
Computer
Science
and
Information
T
echnology
,
Noida
Institute
of
Engineering
and
T
echnology
,
Greater
Noida,
India
Article
Inf
o
Article
history:
Recei
v
ed
Jan
28,
2025
Re
vised
May
13,
2026
Accepted
May
24,
2026
K
eyw
ords:
Deep
learning
License
plate
recognition
Object
recognition
Smart
parking
system
Y
ou
only
look
once
v
ersion
5
ABSTRA
CT
The
inef
ciencies
of
traditional
parking
systems,
including
manual
entry
and
reliance
on
sensors,
as
well
as
slo
w
payment
lines,
contrib
ute
to
congestion
and
a
lo
wer
le
v
el
of
user
satisf
action.
This
paper
suggests
a
smart
parking
management
system
(SPMS)
based
on
the
license
plate
recognition
(LPR)
technology
using
you
only
look
once
v
ersion
5
(Y
OLOv5)
model
in
combination
with
a
QR
code
based
payment
system,
to
automate
the
parking
system.
It’
s
able
to
detect
v
ehicle
license
plates
in
real-time,
which
means
that
data
doesn’
t
ha
v
e
to
be
entered
manually
,
and
it
also
a
v
oids
the
need
for
ph
ysical
sensors,
which
cuts
do
wn
on
infrastructure
costs.
W
ith
a
centralized
database,
the
ef
cient
tracki
ng
of
v
ehicles
is
achie
v
ed,
and
QR-based
payment
will
f
acilitate
contactless
payments
at
e
xit.
Experimental
e
v
aluation
sho
ws
that
the
proposed
system
out
performs
than
the
other
systems,
with
an
accurac
y
of
98.09%
and
recall
of
98.25%
in
case
of
LPR.
This
system
drastically
decreases
processing
time,
c
ongestion
and
impro
v
es
o
v
erall
user
e
xperience.
SPMS
is
more
ef
cient,
scalable,
and
reliable
than
tradit
ional
and
e
xisting
smart
parking
systems.
The
results
sho
w
that
the
proposed
method
is
a
cost-ef
fecti
v
e
and
practical
solution
for
solving
the
modern
parking
management
problem.
This
is
an
open
access
article
under
the
CC
BY
-SA
license
.
Corresponding
A
uthor:
Sumit
K
umar
Department
of
Articial
Intelligence
and
Machine
Learning,
Symbiosis
Institute
of
T
echnology
Pune
Symbiosis
International
(Deemed
Uni
v
ersity)
Pune
412115,
Maharashtra,
India
Email:
er
.sumitkumar21@gmail.com
1.
INTR
ODUCTION
As
urban
areas
e
xpand
and
v
ehicle
o
wnership
rises,
parking
management
has
become
a
signicant
challenge
for
city
planners
and
residents.
T
raditional
systems
of
parking
that
use
manual
entry
for
license
plate
numbers
and
older
payment
methods
ha
v
e
reached
a
stage
where
the
y
are
becoming
completely
unrealistic
for
modernized
urban
centers.
These
systems
are
prone
to
errors,
re
v
enue
losses,
and
inef
ciencies
in
operations.
Manual
processes
in
recording
car
information
tend
to
be
error
-prone,
and
traditional
payment
methods
are
usually
slo
w
and
incon
v
enient
[1],
[2].
Recent
technological
adv
ancements
introduce
promising
alternati
v
es
to
traditional
approaches.
Automation,
articial
intelligence
(AI),
and
real-time
data
processing
can
transform
parking
management
because
the
y
enhance
the
f
acilities’
accurac
y
,
ef
cienc
y
,
and
con
v
enience
[3],
[4].
F
or
e
xample,
li
cense
plate
recognition
(LPR)
technology
is
an
automatic
number
-pla
te
recognition
system
that
has
J
ournal
homepage:
http://ijai.iaescor
e
.com
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Artif
Intell
ISSN:
2252-8938
❒
3615
been
re
v
ealed
to
be
a
promising
application
that
assists
in
tracing
and
tracking
v
ehicles,
which
reads
license
plates
accurately
and
quickly
without
human
interv
ention
[5],
[6].
In
addition,
the
de
v
elopment
of
payment
technologies,
such
as
QR
codes,
has
made
the
check
out
process
much
more
ef
c
ient,
with
contactless
payment
options,
minimizing
w
ai
ting
time
and
better
satisf
action
to
w
ards
usage
[7],
[8].
Such
inte
gration
of
these
technologies
into
a
unied
automated
parking
solution
presents
a
big
step
forw
ard
in
o
v
ercoming
the
e
xisting
limitations
of
traditional
parking
systems.
Adv
anced
algorithms
and
automation
create
more
possibilities
for
ef
cienc
y
in
parki
n
g,
reduced
operational
costs,
and
seamless
e
xperience
for
users
and
operators
alik
e
[9],
[10].
Such
technology-dri
v
en
solutions
are
essential
for
modernizing
current
urban
parking
infrastructure
to
adapt
to
the
needs
of
e
v
er
-changing
city
en
vironments.
One
of
the
major
problems
with
con
v
entional
parking
management
systems,
which
depend
on
hands-on
mechanics
and
outdated
technology
,
is
that
license
plate
numbers
are
entered
manually
in
these
mechanisms,
making
it
erroneous
and
increasing
inaccurac
y
in
tracking
the
v
ehicle
leading
to
re
v
enue
leakage
[11],
[12].
Inef
cient
payment
methods,
such
as
cash
and
card
terminals,
create
long
w
ait
times
that
users
do
not
appreciate
[13].
Furthermore,
the
ph
ysical
sensors
and
the
comple
xity
of
this
infrastructure
increase
the
costs
of
maintenance
and
hampers
system
management
[14].
It
contrib
utes
as
much
to
reduce
operational
ef
cienc
y
as
it
lo
wering
the
quality
of
the
genera
l
parking
e
xperience
in
urban
areas.
As
cities
e
xpand,
the
numbers
mount,
and
as
v
ehicle
re
gistrations
gro
w
,
the
insurm
ountable
limitations
of
such
systems
be
gin
to
pre
v
ail
better
and
assume
a
ne
w
solution
ur
gently
.
Moreo
v
er
,
supporting
these
decits
is
necessary
for
urban
parking
management
to
a
v
oid
congestion,
loss
of
re
v
enues,
and
user
incon
v
enience.
Therefore,
some
inno
v
ati
v
e
approaches
in
v
olving
technology
are
required
[15],
[16].
T
o
o
v
ercome
the
dra
wbacks
of
the
traditional
parking
management
system,
the
automated
parking
solution
proposes
the
inte
gration
of
adv
anced
technologies
to
impro
v
e
ef
cienc
y
,
accurac
y
,
and
user
con
v
enience.
The
proposed
system
has
automatic
license
plate
recognition
(ALPR)
using
the
you
only
look
once
v
ersion
5
(Y
OLOv5)
model
to
capture
the
number
of
license
plates
with
a
high
le
v
el
of
precision
at
the
entry
point
for
each
v
ehicle.
This
implies
that
there
is
no
pro
vision
for
manual
data
entry
in
this
automated
system,
thus
eliminating
error
possibilities
and
ensuring
accurate
billing
[17],
[18].
The
system
emplo
ys
a
sensor
-free
infrastructure,
capturing
only
the
plate
number
from
the
v
ehicle
to
track,
thus
simplifying
the
system’
s
architecture
and
reducing
maintenance
costs
[19].
T
o
enhance
the
user
e
xperience
e
v
en
further
,
QR
code
payment
has
also
been
incorporated
into
the
system
at
the
e
xit,
which
mak
es
payment
e
v
en
simpler
and
reduces
w
aiting
time
at
booths
[20],
[21].
All
these
technologies
are
inte
grated,
allo
wing
an
optimal
use
of
the
space,
and
making
operat
ions
more
ef
cient
and
pleasant
for
users
and
operators.
The
simulations
sho
wed
that
the
system
w
as
able
to
ef
fecti
v
ely
alle
viate
congestion
and
impro
v
e
re
v
enues
[22],
[23].
This
ne
w
solution
can
thus
be
used
to
successfully
solv
e
the
ne
w
challenges
of
parking:
parking
f
acilities
in
the
cities
of
the
future.
The
main
goal
of
the
proposed
smart
parking
management
system
(SPMS)
is
to
create
an
intel
ligent
and
automated
approach
to
impro
v
e
the
ef
cienc
y
,
accurac
y
,
and
user
e
xperience
in
today’
s
parking
landscapes.
It
uses
state-of-the-art
technology
,
including
Y
OLOv5-based
ALPR,
for
rapid
and
accurate
v
ehicle
recognition
at
access
points,
eliminating
manual
data
entry
and
mitig
ating
human
error
.
It
is
an
infrastructure
that
does
not
require
an
y
sensor
to
use
the
v
ehicle
license
plate
information
to
track
v
ehicles,
which
reduces
the
comple
xity
of
the
system
and
lo
wers
maintenance
costs.
Besides,
QR
c
od
e
payment
will
be
incorporated
at
the
e
xit,
which
will
create
a
seamless,
contactless,
and
ef
cient
payment
process,
a
v
oiding
the
need
to
ha
v
e
payment
stations
and
reduce
payment
time.
In
addition,
the
proposed
system
is
designed
for
space
optimization
and
ef
cient
o
v
erall
operations,
which
will
allo
w
the
system
to
manage
parking
resources
ef
fecti
v
ely
and
contrib
ute
to
an
enhanced
user
e
xperience,
both
for
the
operators
and
v
ehicle
o
wners.
The
remainder
of
this
paper
is
or
g
anized
as
follo
ws.
Section
2
deals
with
the
state
of
the
art
of
parking
techniques.
Section
3
describes
the
methodology
of
the
SPMS
m
od
e
l.
Section
4
presents
and
discusses
the
results.
Finally
,
section
5
concludes
the
paper
.
2.
LITERA
TURE
REVIEW
W
ith
the
de
v
elopment
of
smart
parking
management,
man
y
inno
v
ati
v
e
solutions
ha
v
e
been
proposed
to
solv
e
urban
parking
problems.
These
approaches
utilize
adv
anced
technologies
such
as
internet
of
things
(IoT),
deep
learning,
and
automated
systems
to
enhance
parking
solutions’
ef
cienc
y
,
con
v
enience,
and
sustainability
.
Here
is
a
comparison
of
some
note
w
orth
y
rec
ent
literature
contrib
utions
a
v
ailable
in
T
able
1.
It
of
fers
a
clear
Smart
parking
mana
g
ement
system:
a
seamless
parking
solution
using
Y
OLO
and
QR
...
(Sumit
K
umar)
Evaluation Warning : The document was created with Spire.PDF for Python.
3616
❒
ISSN:
2252-8938
e
xample
of
the
dif
ferent
technologies
and
model
s
used,
of
the
main
characteristics
of
each
study
,
and
of
the
main
contrib
utions
of
each,
thereby
highlighting
the
state
of
research
and
the
di
v
ersity
of
strate
gies
used
to
impro
v
e
parking
management
systems.
T
able
1.
Literature
re
vie
w
of
parking
management
systems
Ref.
Model
used
Characteristics
Adv
antages
[1]
Backcasting
approach
Reuse
and
reallocation
of
parking
and
roadspace
Polic
y
measures
for
autonomous
v
ehicle
(A
V)
implementation
and
urban
planning
[2]
Android
app
with
IoT
module
Real-time
parking
space
notication
and
reserv
ation
Lo
w-cost
enterprise
solution
for
smart
parking
[3]
Dynamic
and
proacti
v
e
planning
system
Sensor
technology
and
automated
carriers
Impro
v
ed
operational
ef
cienc
y
via
simulation–optimization
[4]
V
arious
sensors
and
data
analysis
Occupanc
y
detection
and
dynamic
pricing
Challenges
and
opportunities
in
implementing
smart
parking
[5]
V
arious
inno
v
ations
and
technologies
Unb
undling
parking
costs
and
reducing
minimums
Identies
and
outlines
v
e
k
e
y
de
v
elopments
in
parking
[6]
W
illingness
to
pay
(WTP)
model
Hypothetical
permit
program
and
double
hurdle
model
Estimates
residents’
WTP
for
parking
permits
[7]
Open
source
computer
vision
library
(OpenCV),
nump
y
library
and
ArUco
mark
ers
Measurement
of
v
ehicle
size
and
real-time
detection
Ef
cient
space
allocation
and
v
ehicle
monitoring
[8]
Deep
learning-based
object
detection
LPR
and
disability
badge
v
erication
Ef
cient
and
real-time
management
of
accessible
parking
spaces
[9]
Con
v
olutional
neural
netw
orks
(CNN)
Image
classication,
localization,
and
detection
Role
of
deep
learning
in
visual
recognition
systems
[10]
Y
OLOv5
for
v
ehicle
detection
Real-time
performance
and
high
accurac
y
Scalable
and
ef
cient
data-dri
v
en
parking
solution
[11]
Y
OLO-based
model
Real-time
detection
and
number
plate
recognition
High
accurac
y
in
v
ehicle
and
plate
detection
[12]
Y
OLOv5l,
con
v
olutional
recurrent
neur
al
netw
ork
(CRNN)
and
adapti
v
e
feature
fusion
netw
ork
(AFF-Net)
High
accurac
y
and
cloud
deplo
yment
Rob
ust
LPR
system
for
outdoor
parking
toll
[13]
Y
OLOv5
neural
netw
orks
Real-time
v
ehicle
detection
and
f
alse
count
pre
v
ention
Ef
cient
real-time
v
ehicle
video
detection
[14]
Raspberry
Pi
and
deep
learning
tools
Real-time
a
v
ailability
and
automatic
billing
Cost-ef
fecti
v
e
and
time-sa
ving
smart
parking
solution
[15]
IoT
and
v
arious
smart
applications
Adv
ance
reserv
ation
and
cashless
billing
Comprehensi
v
e
features
for
smart
city
inte
gration
3.
METHOD
The
SPMS
is
designed
with
an
emphasis
on
automation,
ef
cienc
y
,
and
user
con
v
enience.
The
system
architecture
is
di
vided
into
four
modules:
admin
panel,
LPR,
user
entry
,
and
user
e
xit.
These
modules
w
ork
in
tandem
to
pro
vide
a
seamless
parking
e
xperience.
The
core
components
of
this
model
include
the
Y
OLOv5
[22]–[26]
model
for
ALPR,
a
centralized
databas
e
for
v
ehicle
tracking,
and
a
QR
code-based
payment
system.
Figure
1
demonstrates
the
SPMS.
The
admin
panel
manages
parking
lots,
congures
fees,
and
tracks
space
a
v
ailability
in
real-time.
LPR
automates
v
ehicle
license
plate
number
e
xtraction
and
database
entries.
QR
code-based
payment
mak
es
it
con
v
enient
for
users
to
mak
e
payments.
The
system
completes
the
payment
using
a
QR
code
when
the
user
lea
v
es.
This
process
mak
es
controlling
traf
c
mo
v
ements
in
parking
zones
easier
and
impro
v
es
operational
ef
cienc
y
.
In
this
module,
a
good
quality
camera
is
used
to
capture
the
v
ehicle
license
plate
at
the
entrance,
which
guarantee
a
high-resolution
image
carrying
all
essential
details
of
that
image,
which
is
the
foremost
thing
for
image
processing.
Figure
2
demonstrates
the
w
orking
of
the
LPR
module.
After
capturing
the
license
plate
images,
some
preprocessing
techniques
are
applied
to
the
pictures
as
the
captured
images
are
in
the
form
of
red,
green,
and
blue
(RGB),
which
is
con
v
erted
into
grayscale.
Dif
ferent
ltering
techniques
remo
v
e
noise
from
the
pictures
to
impro
v
e
the
picture
quality
.
Images
are
resized
into
a
particular
format
afterw
ard.
Image
resizing
smoothens
processing
by
the
CNN.
CNN
architecture
is
emplo
yed
to
identify
license
plate
r
e
gi
o
ns
.
It
e
xtracts
rele
v
ant
features
specic
to
the
license
plates.
Pooling
layers
reduce
the
dimensionality
of
e
xtracted
feature
maps
and
the
rectied
linear
unit
(ReLU)
acti
v
ation
function
is
used
to
learn
comple
x
patterns
in
images.
Dropout
layers
are
used
to
reduce
Int
J
Artif
Intell,
V
ol.
15,
No.
4,
August
2026:
3614–3624
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Artif
Intell
ISSN:
2252-8938
❒
3617
the
o
v
ertting
of
the
model.
Once
the
license
plate
has
been
detected,
indi
vidual
characters
are
se
gmented
from
that
plate
re
gion.
The
se
gmented
characters
are
then
passed
through
an
optical
character
recognition
(OCR)
engine
that
con
v
erts
them
to
alphanumeric
te
xt.
The
e
xact
license
plate
number
is
the
system’
s
nal
output.
The
e
xtracted
te
xt
and
alphanumeric
characters
automatically
get
written
into
the
database
along
with
the
timestamp
of
the
user
who
made
the
request
and
selected
block
information.
This
process
is
automated,
eradicating
the
need
to
manually
input
data
and
ensuring
the
v
ehicle’
s
entry
and
e
xit
times
timestamping.
Figure
1.
SPMS
methodology
Figure
2.
LPR
modules
As
soon
as
the
user
steps
into
the
parking
area,
LPR
is
automatically
in
v
ok
ed,
wherein
the
number
plate
will
be
identi
ed
and
the
number
pick
ed
up.
F
ollo
wing
this,
the
user
will
be
ask
ed
to
choose
a
parking
Smart
parking
mana
g
ement
system:
a
seamless
parking
solution
using
Y
OLO
and
QR
...
(Sumit
K
umar)
Evaluation Warning : The document was created with Spire.PDF for Python.
3618
❒
ISSN:
2252-8938
block.
Three
attempts
are
gi
v
en
to
the
user
to
arri
v
e
at
the
correct
block.
F
or
e
v
ery
attempt
performed,
an
alternati
v
e
block
is
returned.
If
the
user
refuses
the
three
options,
the
block
automatically
gets
assigned.
After
determining
the
block,
it
inserts
the
database
entry
and
its
timestamp.
The
occupied
and
a
v
ailable
blocks
are
displayed
on
the
screen
for
the
user
.
At
the
e
xit,
the
dynamic
QR
code
is
displayed
on
the
light
emitting
diode
(LED)
screen.
The
parking
fee
is
calculated
based
on
the
period
a
v
ehicle
has
been
k
ept
in
the
parking
by
computing
the
dif
ference
in
timestamp
v
alues
from
the
entry
to
e
xit
time
stamp.
Although
the
parking
fee
w
as
computed
hourly
,
the
payment
can
be
managed
with
the
help
of
the
abo
v
e-mentioned
QR
code
through
v
ery
rapid
contactless
payments
without
an
y
cash
handling.
Once
payment
is
conrmed,
the
barrier
opens
and
the
v
ehicle
is
free
to
lea
v
e.
The
Y
OLO
architecture,
as
sho
wn
in
Figure
3,
is
b
uil
t
around
a
sequence
of
con
v
olutional
layers
that
progressi
v
ely
e
xtract
visual
features,
supported
by
residual
connections
that
help
retain
information
across
deeper
layers.
Max
pooling
reduces
spatial
comple
xity
,
while
concatenation
layers
mer
ge
feature
maps
from
dif
ferent
stages
to
preserv
e
both
ne-grained
and
high-le
v
el
details.
Figure
4
presents
the
use
case
diagram
for
the
SPMS,
outlining
the
k
e
y
interactions
between
users,
v
ehicles,
and
the
system
across
entry
,
payment,
and
e
xit
scenarios.
Figure
3.
Y
OLO
architecture
Figure
4.
Use
case
diagram
There
a
re
tw
o
users
in
the
system:
the
admin
and
the
user
.
The
admin
manages
and
congures
the
system;
therefore,
it
has
all
the
needed
permissions.
The
’enter
parking
lot’
acti
vity
describes
the
o
w
of
entering
a
parking
lot.
In
t
his
acti
vity
,
the
user
chooses
a
parking
block
by
selecting
one,
then
goes
and
looks
for
his
v
ehicle
in
a
gi
v
en
spot
by
scanning
it
in.
The
’e
xit
parking
lot’
acti
vity
describes
the
user’
s
e
xit
process.
The
user
scans
a
dynamic
QR
code
to
settle
his
payment
before
lea
ving
the
parking
lot.
The
receipt
is
generated
after
the
e
xit
and
a
QR
code
for
the
payment
are
issued.
After
scanning
t
h
e
code
at
the
payment,
the
v
ehicle
is
released
from
the
parking
lot.
Such
an
automated
system
will
sa
v
e
managers
and
car
o
wners
much
time
by
cutting
of
f
human
interference
in
managing
permits.
The
system
tracks
v
ehicles
using
ALPR
technology
Int
J
Artif
Intell,
V
ol.
15,
No.
4,
August
2026:
3614–3624
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Artif
Intell
ISSN:
2252-8938
❒
3619
without
a
sensor
.
By
doing
this,
ground
loops
and
other
con
v
enti
onal
parking
sensors—prone
to
wear
and
tear
and
necessitate
frequent
maintenance
and
replacement
costs—are
no
longer
necessary
.
The
SPMS
system
simplies
infrastructure
setup
and
reduces
long-term
maintenance
requirements
by
using
LPR
data
for
v
ehicle
identication
and
tracking
throughout
the
parking
duration.
This
approach
enhances
user
con
v
enience
and
eliminates
errors
associated
with
manual
data
entry
,
impro
ving
data
accurac
y
for
billing
and
tracking
purposes.
T
ar
get
detection
using
the
Y
OLO
model,
the
system
identies
the
v
ehicle
at
the
entrance
by
using
Y
OLO
for
real-time
object
detection.
The
single-stage
approach
mak
es
Y
OLO
more
computationally
ef
cient
and
allo
ws
the
video
stream
from
the
camera
at
the
entrance
to
be
processed
in
real-time.
Once
the
user
e
xits
the
parking
lot,
a
QR
code
will
appear
on
the
screen.
Users
automatically
connect
to
a
secure
payment
site
by
scanning
a
QR
code
with
their
smartphone.
This
eliminates
traditional
payments
and
w
aiting
times.
The
QR
code
payment
system
also
simplies
the
payment
process
for
emplo
yees
and
reduces
the
comple
xity
of
managing
cash
or
ph
ysical
payment
f
acilities.
Figure
5
sho
ws
the
o
v
erall
w
orko
w
of
the
system.
The
diagram
sho
ws
the
detection
of
the
li
cense
plate;
if
the
license
plate
is
not
detected,
a
repeti
tion
of
t
h
e
proces
s
i
s
perf
o
r
med,
follo
wed
by
inserting
the
detected
license
plate
in
the
database,
further
calculating
the
time
of
the
v
ehicle
as
the
user
selects
the
slot.
The
main
interf
ace
of
the
v
ehicle
parking
system
gi
v
es
a
li
v
e
vie
w
of
all
parking
slots,
clearly
sho
wing
which
spots
are
tak
en
and
which
are
free.
Free
slots
are
highlighted
in
green,
while
occupied
ones
appear
in
red.
The
panel
on
the
left
w
alks
through
the
v
ehicle
entry
process.
The
user
selects
a
car
image,
and
the
LPR
module,
po
wered
by
Y
OLO,
automatically
reads
the
plate
number
and
sa
v
es
it
to
the
system
database.
The
right-side
windo
w
demonstrates
the
e
xit
process,
displaying
the
current
v
ehicles
par
k
ed.
Upon
clicking
the
e
xit
b
utton,
a
QR
code
is
generated.
When
a
user
scans
the
QR
code,
a
summary
,
including
the
payable
amount,
is
pro
vided
to
the
user
.
Figure
5.
W
orko
w
diagram
of
the
system
Smart
parking
mana
g
ement
system:
a
seamless
parking
solution
using
Y
OLO
and
QR
...
(Sumit
K
umar)
Evaluation Warning : The document was created with Spire.PDF for Python.
3620
❒
ISSN:
2252-8938
4.
RESUL
TS
AND
DISCUSSION
The
performance
of
the
proposed
SPMS
meets
and
e
v
en
outperforms
in
certain
aspects
the
pre
vious
studies
related
to
intelligent
parking
and
LPR.
Recently
,
deep
learning-based
LPR
methods,
such
as
Y
OLO
v
ariants
and
CNN-based
methods
[9],
ha
v
e
achie
v
ed
high
accurac
y
of
95%–98%
based
on
the
comple
xity
of
the
dataset
and
the
en
vironment
in
which
the
y
are
applied.
F
or
e
xample,
Y
OLO-based
LPR
models
[10]–[13]
ha
v
e
pro
v
en
to
be
ef
cient
and
rob
ust
for
real-time
license
plate
detection
in
controlled
conditions,
b
ut
their
performance
drops
slightly
in
the
real
w
orld
where
lighting
changes,
occlusion
and
m
otion
blur
e
xist.
In
contrast,
the
accurac
y
and
recall
of
the
proposed
system
are
98.09%
and
98.25%,
respecti
v
ely
,
which
are
competiti
v
e
and
reliable
results
e
v
en
in
practical
considerations.
Moreo
v
er
,
pre
vious
studies
on
smart
parking
systems
using
IoT
and
image
processing
technologies
[15]
ha
v
e
pointed
out
the
benets
of
impro
v
ement
in
the
automation
of
the
parking
system
and
the
reduction
of
congestion,
b
ut
man
y
of
these
systems
are
based
on
sensor
-based
infras
tructures
or
semi-automated
w
ork
o
ws,
adding
to
maintenance
costs
and
system
comple
xity
.
The
proposed
SPMS,
on
the
other
hand,
reduces
the
dependenc
y
on
ph
ysical
sensors,
uses
vision
based
LPR
and
reduces
infrastructure
o
v
erhead
while
maintaining
high
accurac
y
.
Besides,
the
use
of
QR
code
for
payment
has
been
considered
in
some
recent
studies
on
smart
mobility
as
a
w
ay
to
impro
v
e
the
ease
of
use
for
users,
b
ut
in
the
pre
vious
ones
it
is
not
widely
connected
to
capabilities
of
real-time
LPR
and
automatic
slot
allocation.
The
proposed
system
outperforms
the
e
xisting
syst
ems
in
terms
of
pro
viding
a
more
comprehensi
v
e
and
more
scalable
solution,
where
the
Y
OLO-based
LPR,
automated
database
management
and
contactless
QR-based
payment
are
considered.
The
results
of
this
study
further
conrm
that
the
proposed
SPMS
is
not
only
consistent
with
the
state-of-the-art
methodologies
b
ut
also
to
pro
vide
an
impro
v
ement
re
g
arding
operational
ef
cienc
y
,
human
interv
ention
reduction
and
user
e
xperience.
The
performance
comparison
sho
wn
in
Figure
6
demonstrates
that
the
proposed
SPMS
achie
v
es
superior
res
u
l
ts
compared
to
e
xisting
systems.
The
model
attains
an
accurac
y
of
98.09%
and
a
recall
of
98.25%,
which
are
signicantly
higher
than
traditional
approaches.
This
impro
v
ement
is
primarily
attrib
uted
to
the
use
of
the
Y
OLO-based
LPR
module,
which
ensures
ef
cient
and
precise
detection
in
real-time
scenarios.
The
results
v
alidate
the
rob
ustness
and
reliability
of
the
proposed
system.
Figure
6.
Performance
comparison
of
the
proposed
SPMS
with
e
xisting
systems
A
comparati
v
e
analysis
of
bot
h
the
traditional
parking
system
and
the
proposed
SPMS
is
sho
wn
in
Figure
7
with
re
g
ard
to
v
arious
performance
objecti
v
es
dimensions
lik
e
ef
cienc
y
,
accurac
y
,
and
the
satisf
action
of
the
user
.
The
automated
w
orko
w
,
minimal
human
in
v
olv
ement
and
real-time
processing
of
the
SPMS
mak
e
it
a
signicant
impro
v
ement
o
v
er
the
traditional
system.
QR
code
payment
and
ALPR
are
also
inte
grated
into
impro
v
e
the
user
e
xperience
and
reduce
the
w
orkload.
Int
J
Artif
Intell,
V
ol.
15,
No.
4,
August
2026:
3614–3624
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Artif
Intell
ISSN:
2252-8938
❒
3621
The
comparison
process
ing
time
of
the
proposed
SPMS
and
other
traditional
methods
presented
in
Figure
8
points
out
the
ef
cienc
y
of
the
proposed
method
in
minimizing
the
processing
time.
The
system
can
substantially
reduce
the
w
aiting
time
for
v
ehicles
to
enter
,
pay
for
,
and
lea
v
e
parking
spaces
from
the
con
v
entional
parking
mode.
Through
the
use
of
automated
LPR
and
contactless
QR
code
payment,
the
system
eliminates
manual
operations
and
reduces
w
aiting
time,
thus
promoting
a
seamless
and
ef
cient
traf
c
o
w
in
the
parking
area.
Figure
7.
Comparison
of
traditional
parking
system
and
proposed
SPMS
Figure
8.
Processing
time
comparison
between
traditional
parking
system
and
SPMS
Smart
parking
mana
g
ement
system:
a
seamless
parking
solution
using
Y
OLO
and
QR
...
(Sumit
K
umar)
Evaluation Warning : The document was created with Spire.PDF for Python.
3622
❒
ISSN:
2252-8938
In
T
able
2,
the
comparisons
of
the
SPMS
and
the
traditional
parking
management
system
are
sho
wn
with
adv
antages
and
disadv
antages.
From
the
table,
it
is
e
vident
that
t
he
system
outperforms
traditional
systems
in
man
y
respects.
The
ef
fecti
v
eness
of
the
LPR
system
has
been
v
eri
ed
since
the
related
accurac
y
and
recall
rates
are
quite
high.
QR
code
payment
automation
helps
to
complete
the
process
with
fe
wer
human
errors
and
increased
user
satisf
action.
The
syst
em
utilizes
optimal
space
and
generates
more
re
v
enue
through
well-designed
systems.
Ho
we
v
er
,
the
system
has
disadv
antages
such
as
a
higher
initial
cost
and
greater
reliance
on
technology
than
con
v
entional
systems.
Despite
these
issues,
the
all-around
benets
in
ef
cienc
y
,
accurac
y
,
and
user
-friendliness
point
to
a
highly
desirable
system
compared
with
con
v
entional
parking
management
approaches.
T
able
2.
Comparison
of
proposed
SPMS
and
traditional
parking
system
Feature
SPMS
T
raditional
parking
system
License
plate
entry
ALPR
using
Y
OLO
model
Manual
entry
P
ayment
method
QR
code
payment
P
ayment
centers
with
cash
or
card
Ef
cienc
y
High,
streamlined
process
Lo
w
,
manual
processes
Human
error
Minimal,
automated
system
High,
prone
to
mistak
es
Congestion
Reduced
Higher
due
to
manual
operations
User
satisf
action
Increased
Lo
wer
Con
v
enience
High,
due
to
automation
and
QR
code
payment
Lo
w
,
manual
processes
and
payment
centers
Accurac
y
High
V
ariable,
depends
on
manual
input
Initial
cost
High
Lo
w
T
echnology
dependenc
y
High
Lo
w
Space
utilization
Optimized
Less
optimized
Re
v
enue
generation
Increased
due
to
ef
cienc
y
and
optimized
space
use
Standard,
depends
on
usage
Adaptability
and
capacity
b
uilding
Potential
for
high
adaptability
and
capacity
b
uilding
with
further
impro
v
ements
Limited
5.
CONCLUSION
This
study
proposes
a
ne
w
SPMS
that
tak
es
the
concept
of
traditional
parking
to
the
ne
xt
le
v
el
with
adv
anced
technologies.
SPMS
of
fers
ALPR
technology
with
Y
OLO
models,
which
f
acilitates
v
ehicle
tracking
and
billing
while
minimizing
the
comple
xity
of
processes
by
lo
wering
the
need
of
manual
entry
operations.
The
parking
e
xperience
has
been
enhanced,
and
customer
con
v
enience
has
been
impro
v
ed
with
QR
code
payment.
Through
e
xtensi
v
e
simulation
and
real
w
orld
e
x
ecutions,
it
has
been
demonstrated
that
SPMS
can
sa
v
e
mone
y
,
generate
re
v
enue,
and
enhance
customer
satisf
action.
The
e
xploration
also
of
fers
man
y
benets
for
parking
system
optimization,
as
it
of
fers
a
ne
w
f
ashion
and
can
ha
v
e
the
potential
to
be
widely
used
in
ci
vic
conte
xts.
Ov
erall,
the
SPMS
is
a
signicant
adv
ancement
in
parking
solutions,
pro
viding
a
more
ef
cient
and
en
vironmentally
friendly
w
ay
to
address
the
challenge
of
parking
in
modern
cities.
The
results
highlight
the
signicant
changes
that
can
be
achie
v
ed
by
incorporating
state-of-the-art
technologies
into
parking
systems.
The
y
are
als
o
used
to
illustrate
the
w
ay
automated
syst
ems
can
solv
e
the
problems
f
aced
by
traditional
systems
and
to
open
the
w
ay
for
ne
w
inno
v
ations
in
urban
transport
infrastructure.
The
ndings
of
the
study
pro
vide
researchers
with
suggestions
for
i
mpro
vi
ng
the
accurac
y
and
ef
cienc
y
–
something
that
can
be
used
in
smart
city
technologies.
The
decongesti
v
e,
con
v
enient
and
re
v
enue
generating
properties
of
the
SPMS
are
v
ery
benecial
to
the
community
.
The
system
will
re-engineer
parking
and
enhance
user
e
xperience
in
an
y
urban
setting
to
mak
e
it
manageable
and
ef
cient.
In
the
future,
impro
v
ements
might
be
needed
for
the
system
so
that
real-time
data
analytics
and
machi
ne
learning
algorithms
can
be
applied
to
better
f
acilitate
dynamic
parking
management
and
predicti
v
e
analytics.
Additional
research
into
scalability
will
help
to
further
impro
v
e
ef
cienc
y
of
dif
ferent
urban
and
types
of
parking
f
acilities.
Further
impro
v
ement
can
be
made
through
dynamic
pricing
(based
on
demand)
and
adv
anced
analytics
(based
on
users)
that
w
ould
ultimately
increase
scalability
of
the
system.
It
will
help
maintain
the
SPMS
as
dynamic
and
with
respect
to
the
gro
wing
urban
parking
challenges.
FUNDING
INFORMA
TION
No
funding
in
v
olv
ed.
Int
J
Artif
Intell,
V
ol.
15,
No.
4,
August
2026:
3614–3624
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Artif
Intell
ISSN:
2252-8938
❒
3623
A
UTHOR
CONTRIB
UTIONS
ST
A
TEMENT
This
journal
uses
the
Contrib
utor
Roles
T
axonomy
(CRediT)
to
recognize
indi
vidual
author
contrib
utions,
reduce
authorship
disputes,
and
f
acilitate
collaboration.
Name
of
A
uthor
C
M
So
V
a
F
o
I
R
D
O
E
V
i
Su
P
Fu
Sumit
K
umar
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Ruchi
Rani
✓
✓
✓
✓
✓
✓
✓
Sanjee
v
K
umar
Pippal
✓
✓
✓
✓
✓
✓
✓
C
:
C
onceptualization
I
:
I
n
v
estig
ation
V
i
:
V
i
sualization
M
:
M
ethodology
R
:
R
esources
Su
:
Su
pervision
So
:
So
ftw
are
D
:
D
ata
Curation
P
:
P
roject
Administration
V
a
:
V
a
lidation
O
:
Writing
-
O
riginal
Draft
Fu
:
Fu
nding
Acquisition
F
o
:
F
o
rmal
Analysis
E
:
Writing
-
Re
vie
w
&
E
diting
CONFLICT
OF
INTEREST
ST
A
TEMENT
Authors
state
no
conict
of
interest.
D
A
T
A
A
V
AILABILITY
Data
a
v
ailability
is
not
applicable
to
this
paper
as
no
ne
w
data
were
created
or
analyzed
in
this
study
.
REFERENCES
[1]
E.
G.
Gonz
´
alez,
S.
Nogu
´
es,
and
D.
Stead,
“P
arking
futur
es:
preparing
European
cities
for
the
adv
ent
of
automated
v
ehicles,
”
Land
Use
P
olicy
,
v
ol.
91,
Feb
.
2020,
doi:
10.1016/j.landusepol.2019.05.029.
[2]
S.
Ag
arw
al,
S.
Gupta,
P
.
Ag
arw
al,
G.
Sharma,
and
V
.
T
ripathi,
“P
ARKIT
:
an
android-based
real-time
smart
parking
system
using
IoT
,
”
in
2023
10th
International
Confer
ence
on
Computing
for
Sustainable
Global
De
velopment
(INDIA
Com)
,
2023,
pp.
401–406.
[3]
N.
Dudaklı
and
A.
Baykaso
˘
glu,
“
A
simulation–optim
ization-based
planning
and
control
system
for
operations
of
fully
automated
parking
systems,
”
Computer
s
&
Industrial
Engineering
,
v
ol.
189,
Mar
.
2024,
doi:
10.1016/j.cie.2024.109977.
[4]
K.
Askarbek
and
M.
R.
Bahrami,
“Smart
parking
systems:
a
re
vie
w
on
user
interf
ace
features
and
real
w
orld
e
xamples,
”
in
2023
International
Ur
al
Confer
ence
on
Electrical
P
ower
Engineering
(Ur
alCon)
,
Sep.
2023,
pp.
615–622,
doi:
10.1109/UralCon59258.2023.10291071.
[5]
J.
Rosenblum,
A.
W
.
Hudson,
and
E.
Ben
Joseph,
“P
arking
futures:
an
international
re
vie
w
of
trends
and
speculation,
”
Land
Use
P
olicy
,
v
ol.
91,
Feb
.
2020,
doi:
10.1016/j.landusepol.2019.104054.
[6]
Z.
Guo
and
S.
McDonnell,
“Curb
parking
pricing
for
local
residents:
an
e
xploration
in
Ne
w
Y
ork
city
based
on
willingness
to
pay
,
”
T
r
ansport
P
olicy
,
v
ol.
30,
pp.
186–198,
No
v
.
2013,
doi:
10.1016/j.tranpol.2013.09.006.
[7]
S.
V
ashis
htha,
S.
Joshi,
R.
V
erma,
S.
V
erma,
and
V
.
K
umar
,
“Smart
parking
system
using
machine
learning,
”
International
J
ournal
of
Computer
Applications
,
v
ol.
184,
no.
16,
pp.
9–14,
Jun.
2022,
doi:
10.5120/ijca2022922151.
[8]
D.
Neupane,
A.
Bhattarai,
S.
Aryal,
M.
R.
Bouadjenek,
U.
Seok,
and
J.
Seok,
“Shine:
a
deep
learning-based
accessible
parking
management
system,
”
Expert
Systems
with
Applications
,
v
ol.
238,
Mar
.
2024,
doi:
10.1016/j.esw
a.2023.122205.
[9]
A.
R.
P
athak,
M.
P
ande
y
,
and
S.
Rautaray
,
“
Application
of
deep
learning
for
object
detection,
”
Pr
ocedia
Computer
Science
,
v
ol.
132,
pp.
1706–1717,
2018,
doi:
10.1016/j.procs.2018.05.144.
[10]
S.
Raque,
S.
Gul,
K.
Jan,
and
G.
M.
Khan,
“Optimized
real-time
parking
management
frame
w
ork
using
deep
learning,
”
Expert
Systems
with
Applications
,
v
ol.
220,
Jun.
2023,
doi:
10.1016/j.esw
a.2023.119686.
[11]
V
.
Gnanaprakash,
N.
Kanthimathi,
and
N.
Saran
ya,
“
Automatic
number
plate
recognition
using
deep
learning,
”
IOP
Confer
ence
Series:
Materials
Science
and
Engineering
,
v
ol.
1084,
no.
1,
Mar
.
2021,
doi:
10.1088/1757-899X/1084/1/012027.
[12]
Z.
Rao,
D.
Y
ang,
N.
Chen,
and
J.
Liu,
“License
plate
recognition
system
in
unconstrained
scenes
via
a
ne
w
image
cor
rection
scheme
and
impro
v
ed
CRNN,
”
Expert
Systems
with
Applications
,
v
ol.
243,
Jun.
2024,
doi:
10.1016/j.esw
a.2023.122878.
[13]
J.
Mor
gos,
J
.
V
orcak,
and
K.
Hrudkay
,
“P
arking
information
system
with
articial
neural
netw
ork,
”
T
r
ansportation
Resear
c
h
Pr
ocedia
,
v
ol.
74,
pp.
624–631,
2023,
doi:
10.1016/j.trpro.2023.11.190.
[14]
Y
.
Kher
,
A.
Sax
ena,
P
.
S.
T
amizharasan,
and
A.
D.
Joshi,
“Deep
learning-based
smart
parking
management
system
and
b
usiness
model,
”
in
Computer
V
ision
and
Ima
g
e
Pr
ocessing
5th
International
Confer
ence
,
CVIP
2020
,
2021,
pp.
116–127,
doi:
10.1007/978-981-16-1103-2
11.
[15]
G.
N.
Hainalkar
and
M.
S.
V
anjale,
“Smart
parking
system
with
pre
&
post
reserv
ation,
billing
and
traf
c
app,
”
2017
International
Confer
ence
on
Intellig
ent
Computing
and
Contr
ol
Systems
(ICICCS)
,
Madurai,
India,
2017,
pp.
500-505,
doi:
10.1109/ICCONS.2017.8250772.
[16]
T
.
Diw
an,
G.
Anirudh,
and
J.
V
.
T
embhurne,
“Object
detection
using
Y
OLO:
challenges,
architectural
successors,
datasets
and
applications,
”
Multimedia
T
ools
and
Applications
,
v
ol.
82,
no.
6,
pp.
9243–9275,
2023,
doi:
10.1007/s11042-022-13644-y
.
[17]
Y
.
P
anki
v
,
N.
K
unanets
,
O.
Artemenk
o,
N.
V
eretennik
o
v
a,
and
R.
Nebesn
yi,
“Project
of
an
intelligent
recommender
system
for
parking
v
ehicles
in
smart
cities,
”
in
2021
IEEE
16th
International
Confer
ence
on
Computer
Sciences
and
Information
T
ec
hnolo
gies
(CSIT)
,
Sep.
2021,
pp.
419–422,
doi:
10.1109/CSIT52700.2021.9648687.
Smart
parking
mana
g
ement
system:
a
seamless
parking
solution
using
Y
OLO
and
QR
...
(Sumit
K
umar)
Evaluation Warning : The document was created with Spire.PDF for Python.