Indonesian Journal of Electrical
Engineer
ing and Computer Scien
ce
V
o
l. 9, N
o
. 2
,
Febr
u
ar
y 201
8, pp
.
3
3
2
~
3
34
ISSN: 2502-4752, DOI: 10.
11591/ijeecs
.v9.
i
2
.pp332-334
3
32
Jo
urn
a
l
h
o
me
pa
ge
: http://iaescore.c
om/jo
urnals/index.php/ijeecs
Fuzzy C-Means Algorithm Based
Satellite Image Segmentation
Syed Naz
eebur Rehman
1
, Mohameed Ali Huss
ain
2
1
Res
ear
ch s
cho
l
ar,
Inform
ation
Techno
log
y
,
AM
ET Univers
i
t
y
, Che
nn
ai
2
Departm
ent
of
C
om
puter S
cien
ce,
KL Univ
ers
i
t
y
,
Vij
a
yawad
a
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Oct 5, 2017
Rev
i
sed
D
ec 18
, 20
17
Accepte
d Ja
n 2, 2018
In
t
his
paper,
a
n
improved
vers
ion
of
F
uzzy
C
-
M
eans
(FCM)
a
l
gor
ithm
is
proposed
e
fficiently
to
s
egment
t
he
s
atellite
i
mages.
S
egmentat
ion
of
I
mage
is
one
o
f
the
pr
om
is
ing
and
act
ive
res
e
arches
i
n
recen
t
years
.
As
literatur
e
prove
t
h
a
t
r
e
gio
n
s
egm
e
ntat
ion
will
produce
b
e
tter
results.
Hum
an
v
is
ual
percep
tion
is
m
ore
eff
ect
ive
tha
n
a
n
y
m
achin
e
vis
i
on
s
y
s
t
em
s
f
o
r
extracting
sem
a
ntic inform
a
tion
from
im
ag
e.
A
F
CM
a
lgorithm
i
s
developed
t
o es
tim
at
e
param
e
ters
o
f
t
h
e
prior
probabi
liti
es
a
nd
lik
elih
ood
probabili
t
ies.
S
o
FCM
algorithm
is
u
sed
for
segmenting
background
and
island
extracti
on
is
done
based
on
pixel
intensity
.
F
inally
P
e
a
k
S
i
g
n
a
l
t
o
N
o
i
s
e
R
a
t
i
o
(
P
SN
R)
i
s
cal
cula
ted
and
it
h
as
bet
ter
res
u
lts
than
oth
e
r.
K
eyw
ords
:
FCM
PSNR
Satellite I
m
a
g
e
Seg
m
e
n
t
atio
n
Copyright ©
201
8 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Syed Nazeebur Rehm
a
n,
R
e
search
Sc
h
o
l
ar, I
n
f
o
r
m
a
t
i
o
n Tec
h
nol
ogy
,
AM
ET
Uni
v
er
sity
,
Ch
enn
a
i.
1.
INTRODUCTION
Im
ag
e
Seg
m
en
tatio
n
h
e
lps
in
e
x
t
racting
th
e
si
m
ilar
featu
r
es
f
ro
m
the
images
b
a
s
ed
o
n
their
col
o
r,
in
ten
s
ity
a
nd
g
roup
s
them
t
o
g
e
th
er
u
sing
v
ariou
s
s
eg
m
e
n
t
atio
n
t
ec
hni
qu
es.
B
e
f
o
re
p
er
f
o
rm
i
ng
se
gm
ent
a
t
i
o
n
on
i
m
a
ges,
s
o
m
et
im
es
i
m
a
ge
d
e
noi
si
ng
m
u
s
t
b
e
do
ne
t
o
re
m
ove
t
he
n
oi
se
s
fr
om
t
he
i
m
a
ges
a
n
d
cert
a
i
n
e
d
g
e
det
ect
ors
l
i
k
e s
obel
,
p
re
wi
t
t
, k
i
r
esh a
r
e em
pl
oy
ed.
2.
BA
C
KGR
OUN
D
Satellite
i
mag
e
s
eg
m
e
n
t
atio
n
u
s
ing
in
teractiv
e
appro
a
ch
t
o
m
u
lt
i-o
b
j
ective
g
e
n
e
tic
f
u
z
zy
c
lu
stering
i
s
e
xpl
ai
ned
i
n
[
1]
.
Thi
s
a
l
g
o
r
i
t
h
m
concu
r
r
e
nt
l
y
f
i
nds
t
he
c
l
u
s
t
e
ri
n
g
s
ol
u
t
i
on
as
w
el
l
as
e
vol
ves
t
h
e
s
e
t
of
leg
a
lity
m
easu
r
es
t
h
a
t
are
to
b
e
op
ti
m
i
zed
a
t
th
e
sa
m
e
t
i
m
e.
H
u
m
a
n
deci
si
on
m
a
ker
i
s
i
nt
eract
s
wi
t
h
m
et
hod
and
best
s
et
i
s
obt
ai
ne
d
usi
n
g
ada
p
t
i
v
e
l
e
a
r
ni
ng
of
v
al
i
d
i
t
y
m
e
asure
wi
th
f
inal
r
esult.
M
oth-flam
e
based
o
p
tim
izat
io
n
fo
r
satellite
i
mag
e
s
eg
m
e
n
t
atio
n
with
m
u
lti
lev
e
l
t
hresh
o
l
d
i
ng
i
s
p
rese
nt
e
d
i
n
[2]
.
M
ul
t
i
l
e
vel
th
resh
o
l
d
i
ng
m
o
t
h
-fla
m
e
o
p
ti
mizatio
n
algo
rith
m
fo
r
m
u
ltilev
e
l
t
hresh
o
l
d
i
n
g
was
de
vel
o
ped
.
M
ost
of
t
he
satellite
i
mag
e
s
are
tested
u
si
n
g
th
is
m
eth
o
d
.
Th
ere
are
five
e
xi
st
i
ng
m
e
tho
d
s
are
c
o
m
p
ared
h
ere
for
solving
t
h
res
hol
di
n
g
p
ro
bl
em
s
li
ke
d
i
ffere
nt
i
a
l
evol
ut
i
on
al
go
ri
t
h
m
,
g
e
n
e
tic
a
lg
o
r
ith
m
,
p
article
s
warm
o
p
timizatio
n
,
artificial
b
ee
co
lon
y
a
lg
orithm
an
d
m
o
th
-fl
a
m
e
o
p
t
i
m
izati
o
n
al
g
orith
m
.
S
atellite
i
m
a
g
e
s
eg
m
e
n
t
atio
n
u
s
ing
di
ffe
re
nt
t
echn
i
ques
i
s
d
i
s
cu
s
s
ed
i
n
[3]
.
T
h
r
esh
o
l
d
i
n
g
Tec
hni
qu
e
,
A
ct
i
v
e
C
ont
ou
rs
a
nd
K
-m
eans
C
l
us
t
e
ri
ng
are
th
e
t
h
ree
tech
n
i
q
u
e
s
used
f
or
s
eg
m
e
n
tin
g
a
satellite
i
ma
g
e
a
nd
e
s
t
i
m
a
ted
t
h
e
best
m
etho
d
wh
en
c
om
pa
r
e
d
t
o
ot
h
er
m
et
hod.
Co
m
p
arativ
e
stu
d
y
a
bou
t
Satellite
i
m
a
g
e
s
eg
m
e
n
t
atio
n
u
s
i
n
g
cl
u
s
tering
a
lg
orithm
s
b
ased
on
per
f
o
r
m
a
nce
o
f
F
uzzy
i
s
desc
ri
be
d
i
n
[
4]
.
C
l
ust
e
ri
n
g
a
p
p
r
o
aches
b
ased
o
n
po
ssib
ilistic
c
m
ean
s,
p
o
s
sib
i
listic
fuzzy
c
m
eans
and
fuzzy-C-Means
is
c
om
pare
d
a
nd
t
h
es
e
algorithm
s
were
t
este
d
wi
th
p
e
r
form
ance
wit
h
m
o
re
n
u
m
b
e
r
o
f
s
atellites.
C
o
m
p
a
rativ
e
stud
y
abou
t
satellite
i
m
ag
e
segmen
tatio
n
u
s
ing
g
e
n
e
tic
a
l
g
orith
m
Evaluation Warning : The document was created with Spire.PDF for Python.
In
d
onesi
a
n
J
E
l
ec En
g &
C
o
m
p
Sci
ISS
N
:
2
5
0
2
-
47
52
Fu
zzy C-Mea
n
s Algo
rithm Based
Sa
tellite Im
ag
e
S
e
g
m
en
t
a
tio
n (S
yed
Nazeebu
r
Rehman
)
33
3
base
d
o
n
d
i
f
fer
e
nt
o
b
j
ect
i
v
e
f
unct
i
o
ns
i
s
p
r
e
s
ent
e
d
i
n
[
5]
.
Di
f
fe
re
nt
o
b
j
ec
tive
fu
nction
is
e
m
p
loy
e
d
fo
r
im
age
seg
m
en
tatio
n
u
s
ing
g
e
n
e
tic
a
lg
orith
m
.
T
sallis,
Otsu
a
nd
K
ap
ur’s
a
re
t
he
t
hree
o
b
ject
i
v
e
f
unct
i
o
ns
c
om
pare
d
b
a
sed
on
g
en
etic
a
lg
o
r
ith
m
f
o
r
o
p
tim
al
m
u
ltilev
e
l
th
resh
old
i
n
g
.
S
atellite
a
n
d
m
e
d
i
cal
i
mag
e
s
eg
m
e
n
t
ation
base
d
o
n
m
ul
t
i
pl
e
ke
rn
el
f
uzz
y
c
-M
eans
al
g
o
ri
t
h
m
wi
t
h
A
LS
m
et
ho
d
is
e
x
p
l
ain
e
d
i
n
[
6
]
.
In
itial
co
n
t
o
u
r
curv
e
is
g
en
erated
u
sin
g
m
u
ltip
le
k
ern
e
l
fu
zzy
c
-m
ean
s d
u
ring
t
h
e
c
u
r
v
e
p
ro
p
a
g
a
tio
n
wh
ile
l
eakin
g
at
t
h
e
b
ound
ar
y.
Fin
a
lly
d
ifferen
t
i
n
f
o
r
m
a
tio
n
’
s
are
co
m
b
in
ed
u
sing
m
u
ltip
le
k
er
nel
fuzzy
c
-m
eans
in
s
e
g
m
e
ntation
alg
o
rith
m
.
I
m
a
g
e
su
p
e
r
reso
lu
tio
n
recon
s
tru
c
tio
n
using
iterativ
e
a
dapt
i
v
e
reg
u
l
a
ri
zat
i
o
n
m
e
t
hod
a
n
d
genet
i
c
algorithm
explained
i
n
[
7].
Medi
a
Access
Delay
and
T
h
roughput
A
nalysis
o
f
V
o
i
ce
C
o
d
e
c
with
S
ilen
ce
Su
pp
ressi
on
o
n
W
i
r
el
ess
Ad
H
oc
N
et
wo
rk
a
l
s
o
descri
be
d
i
n
[
8]
.
A
n
i
n
tegrated
i
nteractive
techni
que
for
im
age
segm
ent
a
t
i
on
usi
n
g
st
ack
b
ase
d
s
ee
d
e
d
regi
on
g
r
o
wi
n
g
a
nd
t
h
res
hol
di
n
g
m
et
ho
d
i
s
d
i
s
c
u
sse
d
i
n
[
9]
.
For
a
n
alyzing
the
optim
al
p
erfo
rm
ance
of
p
est
im
age
segm
entatio
n
is
d
iscusse
d
as
i
n
[10].
Im
age
segm
entation
base
d
on
d
oubly
tr
uncate
d
g
ene
r
alized
L
aplace
mixt
ur
e
m
odel
and
k
m
eans
cl
ust
e
ri
n
g
i
s
di
scuss
e
d
i
n
[
1
1
]
.
3.
THE PROBLEM
So
m
e
o
f
th
e
p
r
ob
lem
s
i
n
satellite
i
m
a
g
e
s
e
g
m
e
n
t
atio
n
syste
m
a
re
t
h
e
s
e
g
m
e
n
t
atio
n
o
f
t
h
e
s
atellit
e
i
m
ag
e
m
a
y
v
a
ry
w
ith
t
h
e
i
n
t
en
sity
v
ariatio
ns
i
n
t
h
e
satellit
e
i
m
a
g
e
.
So
i
n
ord
e
r
t
o
o
v
e
rco
m
e
th
e
in
tensities
vari
at
i
o
ns i
n t
h
e segm
ent
e
d i
m
ag
e we a
re proposi
ng a new satell
i
t
e
im
a
ge segm
e
nt
at
i
on
m
e
t
hod.
4.
PROP
OSE
D
S
OLUTI
O
N
In
t
h
i
s
p
r
op
osed
s
ystem
,
s
atel
lite
i
mag
e
s
are
seg
m
en
ted
u
s
i
n
g
F
CM
a
lg
orith
m
.
B
efo
r
e
seg
m
en
tatio
n
p
r
e
pro
cessing
i
s
d
o
n
e
u
s
ing
med
i
an
f
ilter
to
d
en
o
i
se
a
n
imag
e
to
g
et
b
etter
resu
lts.
Here,
first
th
e
imag
es
a
re
segm
ented
usi
n
g
this
a
l
g
orithm
.
F
CM
p
arameters
are
use
d
t
o
dete
rm
in
e
th
e
laten
t
v
ariab
l
e
d
i
stri
b
u
tion.
B
lo
ck
di
ag
ram
of t
he
p
r
o
pose
d
i
m
a
ge segm
e
nt
at
i
o
n
i
s
gi
v
en
bel
ow
.
FCM
is
a
m
ethod
o
f
cl
ustering
w
hic
h
a
llows
o
ne
p
iece
of
d
ata
to
b
e
l
o
n
g
t
o
t
w
o
o
r
m
o
r
e
c
l
u
s
t
e
r
s
.
I
t
is
m
ax
i
m
u
m
u
sed
for
p
a
ttern
r
ecogn
itio
n.
C
lu
stering
or
c
lu
ster
a
n
al
y
s
i
s
i
nv
ol
ve
s
assi
gni
ng
d
at
a
p
o
i
n
t
s
t
o
cl
ust
e
rs
(
al
so
c
al
l
e
d
b
u
cket
s,
b
i
n
s
,
o
r
cl
asse
s),
or
h
om
ogeneo
u
s
classes
,
s
uc
h
that
item
s
i
n
the
sam
e
c
lass
or
clu
s
ter are as similar as p
ossible, wh
ile ite
m
s
b
elon
g
i
n
g
to
d
i
ffe
rent classes
are as di
ssim
i
lar as possible.
Fig
u
re 1
. Blo
c
k
Diagram
o
f
th
e
Pro
po
sed
Satellite I
m
a
g
e
Seg
m
e
n
t
atio
n
5.
RESULTS
A
ND
DI
S
C
U
S
S
I
ON
Th
is
s
tep
d
e
scrib
e
s
th
e
ov
eral
l
resu
lts
o
f
t
h
e
p
r
op
o
s
ed
s
ystem
.
N
o
r
m
a
lly
i
mag
e
s
will
su
ffer
fro
m
th
e
n
o
i
se.
Med
i
an
f
ilter
is
p
ropo
sed
in
t
h
i
s
pap
e
r
to
d
en
o
i
se
t
h
e
i
m
a
ge.
Thi
s
s
ect
i
on
t
e
l
l
s
a
bout
p
r
o
pos
ed
seg
m
en
tin
g
sch
e
m
e
u
sin
g
FCM
is
o
pp
ressed
for
satellite
i
m
a
g
e
s.
Per
f
o
rm
ance
of
t
he
p
r
o
pose
d
s
c
h
em
e
i
s
com
put
ed
b
y
P
S
NR
v
al
ue
. Fi
g
u
r
e
2 s
h
ows
t
h
e
pr
o
pose
d
(
a
)
o
ri
gi
na
l im
age, (
b)
se
g
m
e
nted im
a
ge.
(a)
(b
)
Figu
re
2
.
(a
)
O
ri
ginal im
age an
d
(b
) Se
gm
ented
Im
age usi
n
g
FCM
algorith
m
Evaluation Warning : The document was created with Spire.PDF for Python.
ISS
N
:
2502-
4
752
In
d
onesi
a
n
J
E
l
ec En
g &
C
o
m
p
Sci
, Vol
.
9
,
N
o.
2
,
Fe
br
uar
y
20
1
8
:
3
32 – 334
33
4
6.
CO
NCL
USI
O
N
A
no
vel
ap
proach
t
o
seg
m
en
t
th
e
satellite
i
mag
e
s
were
d
ev
elop
ed
u
sing
F
CM
i
s
p
r
opo
sed
.
F
r
o
m
th
e
i
nvest
i
g
at
i
o
nal
resul
t
s
t
he
i
m
a
ge
s
egm
e
nt
at
i
on
usi
n
g
t
h
e
p
r
o
p
o
s
ed
m
et
ho
d
was
fo
u
nd
t
o
b
e
m
o
re
v
i
s
ual
l
y
te
m
p
tin
g
th
an
o
th
er ex
i
sting
alg
o
ri
t
h
m
s
. Fi
gure
2 s
h
o
w
s t
h
e pr
o
pose
d
se
g
m
e
nt
ed i
m
a
ge usi
n
g
FC
M
al
g
ori
t
h
m
t
echni
q
u
e.
T
he
r
esul
t
s
s
ho
w
t
h
at
F
C
M
a
l
gor
i
t
h
m
m
e
t
hod
i
s
a
v
ery
e
fficien
t
o
p
tim
iza
tio
n
an
d
ob
tain
ed
PSNR
val
u
e
i
s
3
6.
2
3
.
F
ut
u
r
e
sc
o
p
e
o
f
t
hi
s
pa
per
i
s
t
o
use
t
h
e
a
dva
nc
e
d
s
e
g
m
e
nt
at
i
o
n
t
ech
ni
que
t
o
o
b
t
a
i
n
m
ore
accurate res
ult.
REFERE
NC
ES
[1]
A.
M
ukhopadhy
ay
,
“Int
eractiv
e
approach
t
o
m
u
ltiobject
ive
geneti
c
fuz
z
y
c
lustering
for
satelli
te
i
m
a
g
e
segm
entation
,
”
I
EEE Ut
tar Pradesh Sect
ion Int
e
rnational Conf
er
ence on
El
ect
rical, Comput
er and Ele
c
troni
c
s
Engineering
, pp. 630-634, 2016.
[2]
Muangkote
N.,
et a
l
.
,
“
M
ulti
leve
l
thr
e
sholdi
ng
for
sate
lli
te
i
m
a
ge
s
egm
e
ntation
with
m
o
th-flam
e
b
a
s
e
d
optimization,”
I
EEE
1
3
th
Intern
ational Jo
int Co
nference on Computer Scien
ce
and Software
En
gineering
,
pp.
1
-
6
,
2016.
[3]
Deepika
N.
P
.
and
Vishnu
K.,
“D
ifferent
t
echn
i
ques
for
satellit
e
im
age
segm
entation
,
”
IE
EE On
line Int
e
rnationa
l
Conference on
Green Engi
n
eering and Techno
lo
gies,
pp
. 1-6
, 20
15.
[4]
Ganesan
P.,
et a
l
.
,
“Performance
o
f
fuzzy
based
clus
ter
i
ng
algorithms
for
the
seg
m
e
ntation
of
s
atellite
i
mages—
A
com
p
arativ
e
s
t
u
d
y
,”
IEEE S
eventh National C
onference on C
o
m
puting, Communication and
Information Sys
Satellite image segmentation bas
ed on d
iff
erent o
b
jec
tiv
e fun
c
tion
s
using gene
tic algorithm:
a co
mparative
stud
y,
pp. 23-27
, 2015
.
[5]
S
. P
are
,
et
al
.,
In
ternational Conf
erence on Dig
ita
l Signal Processing
, pp
. 730-734
, 2015.
[6]
Yugander
P.,
et al.
,
“
M
ultipl
e
k
e
r
nel
fu
zz
y
C-m
e
ans
algor
ithm
with
A
LS
m
ethod
f
or
sat
e
ll
ite
a
nd
m
edica
l
i
m
a
g
e
segm
entation
,
”
I
EEE Internation
a
l
Conference o
n
Devices, Circu
its and S
y
stems
, pp. 244-248
, 20
12.
[7]
Pa
nda
S
.
S.,
et al.
,
“Image
s
u
p
er
r
esolu
tion
r
econs
truc
tion
us
ing
it
era
tive
ad
apti
v
e
r
egu
l
ar
ization
method
and
genetic algor
ith
m,”
in
Computational In
telligen
ce in
Data M
i
nin
g
,
vol. 2
, pp
. 675
-681, 2015
.
[8]
Shah
R
.
D.
a
nd
S
ingh
S.
K
.,
“Media
A
ccess
Delay
and
Thro
ughput
Anal
y
s
is
o
f
Voice
Code
c
with
S
il
en
c
e
Suppression on Wireless Ad Ho
c
Network
,”
Pro
cedia
Computer
Scien
c
e,
vol. 79
,
pp. 940-947, 20
16.
[9]
H
o
re
S
.,
et al.
,
“An
Integrated
I
nteractiv
e
Techn
i
que
f
or
I
mage
S
egmentation
u
s
i
ng
S
t
ack
b
as
ed
S
eeded
R
egion
Growing
and
Thresholding,”
In
ternational Journal
of Electr
ical and Computer
Engineering,
v
o
l
/
i
s
s
u
e
:
6
(
6
)
,
p
p
.
2773-2778, 201
6.
[10]
Sangari
S
.
and
Saraswad
y
D.,
“A
na
ly
zi
ng
t
h
e
Opti
ma
l
Pe
rforma
n
ce
o
f
Pest
I
mage
S
egmentation
us
ing
Non
Linear
Object
ive
As
s
e
s
s
m
ents
,”
International Journal o
f
Electr
ical and
Computer Engin
eering,
vo
l/issue:
6
(6),
pp.
2789-
2793, 2016
.
[11]
J
y
othirmay
i
T.,
et al.
,
“Image
S
egmentation
Bas
e
d
on
Doubly
Tr
uncated
G
ener
alized
L
ap
lac
e
M
i
x
ture
M
ode
l
an
d
K
Means
Clustering,”
Internatio
nal Journal of Elec
trica
l
and Computer Engineering,
vol/issue:
6
(5),
pp.
2188
,
2016.
Evaluation Warning : The document was created with Spire.PDF for Python.