In
d
o
n
e
sian
Jou
r
n
al of
Ele
c
tr
i
c
a
l
En
g
in
e
erin
g
a
n
d
C
om
pu
t
er S
c
ien
ce
Vol.
14, No.
1, April 2019,
pp.
250~257
ISSN: 2502-
4752,
DOI
:
10.115
91/ijeecs.
v
14.
i
1
.
pp250-257
250
Jou
rn
a
l
h
o
me
pa
ge
:
ht
tp:
//i
a
e
score
.
com
/
j
o
u
r
na
l
s
/
i
n
d
e
x
.
p
hp/
i
j
eec
s
Spati
a
l domain image e
nh
ancement techniques for
acute m
y
eloid leukemia (M1,M4,M5,M7)
A.
S
.
A.Sa
l
a
m
1
,
M
.
N.M.Isa
2
, M.
I
.
A
h
m
ad
3
1,
2
S
c
ho
ol
o
f
Mi
croelect
ron
i
c
En
g
i
neeri
ng,
Un
i
v
e
rsit
i
Ma
l
a
ys
ia P
erl
is
,
M
a
l
a
y
s
i
a
3
S
c
ho
ol
o
f
Co
mpu
t
er
a
nd
C
ommu
ni
ca
t
i
o
n
E
ngin
e
ering
,
U
ni
versi
t
i M
a
lay
s
i
a
P
erli
s, M
al
ays
i
a
Art
i
cl
e In
fo
ABSTRACT
A
r
tic
le hist
o
r
y
:
R
e
c
e
i
v
e
d
Sep
3
0
,
2
018
Re
vise
d N
ov
26,
201
8
A
c
c
e
pte
d
N
ov 5,
201
8
In
t
his
pap
e
r,
s
ev
eral
t
ech
ni
que
s
of
i
m
a
g
e
e
nhan
cem
ent
s
p
at
ia
l
do
ma
in
i
s
elu
c
id
ated
a
nd
a
n
a
ly
zed
f
or
t
h
e
p
urpo
se
o
f
enh
a
ncin
g
Acu
t
e
M
y
el
oid
L
e
u
k
e
m
i
a
(
A
M
L
)
s
u
b
t
y
p
e
o
f
M
1
,
M
4
,
M
5
a
n
d
M
7
.
T
h
e
t
e
c
h
n
i
q
u
e
s
i
n
v
olved
con
t
ras
t
s
tretchin
g
o
f
g
rey
s
cale
i
m
ages,
im
a
g
e
su
btract
io
n
an
d
i
m
ag
e
sh
arpen
i
ng
.
Th
e
t
h
ree
m
e
th
od
s
com
p
are
d
w
i
t
h
o
n
e
ano
t
h
e
r
t
o
achi
eve
th
e
hi
gh
est
P
S
NR
v
al
ue
f
o
r
t
he
s
uita
b
ilit
y
t
echn
i
q
u
e
of
A
M
L
s
u
b
ty
pe
s
(M1,
M
4,
M
5
a
n
d
M
7).
F
i
rstly
,
s
u
b
t
yp
es
i
m
a
ges
co
nv
erted
i
n
t
o
g
rayscal
e.
Th
en,
e
ach
f
o
u
r
i
m
a
ges
tested
w
ith
c
o
n
t
r
ast
stre
t
c
hi
ng
t
echn
i
q
u
es
f
o
l
low
e
d
b
y
i
m
a
g
e
su
bt
ractio
n
and
im
ag
e
sh
arpen
i
n
g
.
The
perf
o
r
m
a
n
ces
w
ere
ev
aluat
ed
b
ased
on
M
ean
S
qu
ar
e
E
r
ro
r
(M
S
E
)
and
P
eak
S
ig
nal
t
o
N
o
i
se
R
atio
(
P
S
N
R
).
Du
e
to
its
h
i
g
h
e
r
val
u
e
o
b
t
a
ined
,
i
m
age
s
h
arpen
i
n
g
i
s
a
go
od
en
h
a
ncem
ent
tech
ni
ques
f
o
r
Acu
t
e
M
y
el
oi
d
Leu
k
em
i
a
w
it
h
6
8
.
2
083
d
B
and
th
e
l
ow
es
t
MSE
a
c
h
ie
ve
d of
0
.01
0
3
.
K
eyw
ord
s
:
AML
Co
nt
rast
st
r
et
ch
ing
Im
age
enha
nce
m
ent
Im
age
sharpe
ning
Im
age
subt
r
act
io
n
Leukemia
Co
pyri
gh
t © 2
019 In
stit
u
t
e
of Advanced
En
gi
neeri
n
g
an
d
Scien
ce.
All
rights
res
e
rv
ed.
Corres
pon
d
i
n
g
Au
th
or:
Mo
hd
N
a
z
r
in b
in
M
d I
s
a,
S
c
hoo
l
o
f
Micr
o
elec
tro
n
i
c
En
gine
eri
n
g,
Uni
v
ersi
ti
M
al
ays
i
a P
e
rlis,
P
a
uh P
u
tra
Cam
pus,
0260
0,
A
r
au,
P
e
rlis,
M
a
lay
s
ia.
Em
ail:
nazr
in
@u
n
i
m
a
p.e
du.
m
y
1.
I
N
TR
OD
U
C
TI
O
N
I
n
m
e
d
i
c
i
n
a
l
f
i
e
l
d
,
m
e
d
i
c
a
l
i
m
a
g
i
n
g
i
s
o
n
e
o
f
t
h
e
c
o
n
t
r
o
l
l
i
n
g
a
ppar
a
tus
for
hav
i
ng
a
n
i
ns
ig
ht
o
n
th
e
pat
h
ol
ogic
a
l
p
r
o
ced
ure
s
.
Image
s
t
hat
i
n
c
l
ude
s
c
o
m
pute
r
t
om
ogra
p
hy
(
C
T
),
m
ag
n
e
ti
c
re
so
n
a
n
ce
imag
i
ng
(MRI),
u
lt
r
a
s
o
un
d
a
nd
X
-ray
are
one
o
f
t
h
e
foca
l
base
s
for
d
i
ac
r
i
sis
of
d
ise
a
ses.
T
he
m
a
i
n
pur
p
o
se
s
of
me
dica
l
ima
g
e
proce
s
si
n
g
i
s
to
d
ia
g
nose
m
e
dica
l
im
age
s
m
or
e
eff
ici
e
n
t
ly
a
nd
a
c
c
u
ra
te
ly.
T
y
pica
l
l
y
,
the
s
e
ima
g
es
a
re
a
ffe
c
ted
b
y
n
o
i
se
,
bl
urrines
s
a
n
d
othe
r
ba
d
q
u
a
l
itie
s
t
h
a
t
i
nt
erru
pt
s
t
h
e
qu
al
ity
of
t
h
e
i
ma
g
e
.[
1
]
[2
]
Th
us,
ima
g
e
enha
ncem
en
t
te
c
hni
que
s
c
a
n
i
m
pr
ov
e
t
h
e
v
i
sua
l
a
ppe
a
r
a
nc
e
of
m
edica
l
i
ma
ges
espe
cia
l
l
y
i
n
detec
t
i
ng L
e
u
k
e
m
i
c c
e
ll.
1.1.
Le
u
k
em
i
a
Le
uk
emi
a
a
re
b
on
e
marro
w
c
a
n
c
erou
s
ce
ll
,
wh
i
c
h
i
n
vol
ve
s
p
r
o
l
i
f
e
ra
tio
n
o
f
w
hi
t
e
b
l
o
od
ce
l
l
s
t
h
a
t
di
sa
b
l
e
s
i
ts
m
a
i
n
func
ti
on
t
o
f
i
g
h
t
b
acte
r
ia
m
or
e
effic
i
e
n
t
l
y.
The
b
l
as
t
c
e
l
l
a
r
e
grou
pe
d
by
a
ffec
te
d
blo
o
d
ce
ll
ty
pe
c
a
lled
l
y
m
phoc
y
t
es
a
n
d
m
yel
o
c
y
tes.[3]
T
h
e
d
i
sea
s
e
als
o
c
a
t
ego
r
i
z
ed
b
y
i
t
s
sp
eed
o
f
i
l
l
n
ess
e
s.
Acute
me
ans
the
ce
ll
s
sprea
d
s
fas
t
w
hi
le
c
hron
ic
t
a
k
e
s
t
i
m
e
gro
w
i
n
g
b
u
t
d
o
w
o
r
s
e
n
s
o
v
e
r
t
h
e
y
e
a
r
.
Th
us,
Le
uke
mia
are
gro
u
p
ed
b
y
fo
ur
m
a
i
n
t
ypes
w
h
ic
h
are
acute
l
ym
p
h
o
c
yt
i
c
l
e
ukem
i
a
(
AL
L
),
chro
nic
l
y
mph
o
c
y
tic
l
eu
ke
mi
a
(
CLL
),
a
cute
m
ye
l
o
cy
t
i
c
le
uke
mia
(
AML
)
and
ch
ro
ni
c
my
e
l
o
c
y
t
i
c
l
euk
e
mi
a
(
CML
)
.
A
ML
t
ype
l
e
u
kem
i
a
usu
a
l
l
y
d
ete
c
t
e
d
u
n
t
il
i
t
h
a
s
s
prea
d
i
n
to
oth
e
r
orga
n
s
.
The
ce
lls
c
l
a
ssi
fie
d
b
y
a
sy
st
em
k
no
wn
a
s
Fren
c
h
-A
me
ri
ca
n
B
r
iti
s
h
(FAB
),
w
h
i
ch
c
at
e
g
o
r
i
z
e
d
i
n
to
e
i
g
h
t
s
ub
t
y
pes
a
s
s
how
n
in Ta
b
le
1.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
ones
i
a
n
J
E
lec
En
g & Co
mp
S
c
i
IS
S
N
: 2502-
47
52
Sp
a
tia
l dom
ain
i
m
age
e
nha
nc
em
e
n
t t
e
c
h
n
i
q
u
e
s
for
ac
u
t
e
m
y
elo
i
d leu
k
e
m
ia (M1,
M4,M5,
M7) (A.S.
A
.Sal
a
m
)
25
1
Table
1.
Fr
ench-
A
me
rican-
B
r
i
t
i
s
h
C
l
a
ssific
a
tio
n
[4]
FA
B
Subty
p
e
s
N
am
e
M
0
U
ndiffe
re
nt
i
a
te
d ac
ut
e
m
y
elobla
s
ti
c
l
e
uk
em
i
a
M1
Ac
ut
e
mye
l
obla
s
tic l
e
ukemi
a
with minima
l
ma
t
ura
tion
M2
A
c
ut
e
m
y
e
l
obla
s
tic
l
e
ukem
i
a
w
ith ma
t
ur
a
tion
M3
A
c
ut
e
pr
om
ye
lo
cyt
i
c
l
e
ukem
i
a
(AP
L
)
M4
Acu
t
e
my
el
o
m
o
n
o
cyt
i
c
l
e
u
k
emi
a
M4
e
o
s
A
c
ut
e
m
y
e
l
o
m
ono
c
y
ti
c
l
e
uk
e
m
i
a
with
e
osinophil
i
a
M5
A
c
ut
e
m
ono
c
y
ti
c
l
e
uk
em
i
a
M6
A
c
ut
e
e
r
y
t
hroid
leuke
m
i
a
M7
A
c
ut
e
m
e
g
a
k
a
ry
ob
la
st
i
c
l
e
ukem
i
a
I
n
M
a
l
a
y
sia,
M
y
e
lo
id
t
ype
s
ta
tes
the
hig
h
es
t
am
ou
nt
f
or
b
ot
h
m
a
l
e
a
n
d
f
emal
e
a
s
d
epi
c
t
e
d
in
F
i
gure
1 a
n
d F
i
gur
e 2.
Fi
g
u
r
e 1
.
Le
u
kemi
a t
y
pe
s ag
e-sp
e
c
if
i
c
in
c
i
d
en
ce
ra
t
e
,
m
a
l
es
, M
a
la
ysia,
20
07-
20
1
1
[
5]
F
i
gur
e 2.
Le
uke
m
i
a
typ
e
s a
g
e
-
spe
c
i
f
ic
i
nc
i
d
e
n
ce
rate
, fem
ales,
Malays
ia,
200
7-2
0
1
1
Base
d
o
n
t
h
e
g
r
a
ph,
it
s
h
ow
s
tha
t
M
ye
l
o
i
d
i
n
Ma
la
ys
ia
i
s
s
e
ve
re
.
H
e
nce
,
a
lter
n
a
t
i
v
e
e
a
rl
y
d
i
a
gnos
tic
to
e
n
h
ance
c
urre
nt
hos
p
ita
l
i
t
y
a
r
e
m
uch
nee
d
e
d
.
Th
is
i
s
w
h
e
r
e
m
edica
l
i
m
a
ge
p
roc
e
ss
ing
be
c
o
m
e
s
use
f
u
l
f
or
dia
g
no
sin
g
ear
l
y
d
etec
ti
on o
f
A
ML sub
t
y
pes
.
1.2.
Im
age En
h
an
c
e
me
nt
In
i
m
a
ge
p
roc
e
ssing,
i
m
a
ge
e
nha
nc
em
en
t
i
s
c
a
p
a
b
le
o
f
i
m
prov
ing
the
im
age’
s
co
ntra
st
f
or
m
a
k
in
g
vari
ous
f
e
a
t
u
re
s
to
b
e
e
a
s
ily
r
ec
o
g
n
i
ze
d.[
6
].The
ro
le
i
s
vit
a
l
i
n
e
nha
nc
in
g
q
u
a
lit
y
o
f
m
e
d
ic
a
l
i
ma
ges
suc
h
a
s
AM
L
b
e
c
a
u
s
e
t
h
e
i
m
ag
e
’
s
q
u
a
li
t
y
d
ep
en
ds
o
n
th
e
e
xposu
r
e
o
f
t
h
e
mic
r
osc
ope
a
nd
s
t
a
ini
n
g
proc
e
s
s[7].
En
hance
m
e
n
t t
echn
i
que
s
are
di
v
i
de
d i
n
t
o
tw
o
broa
d
ca
t
eg
o
r
ies:
a)
Spa
t
ia
l
d
o
ma
in
m
e
t
hod
s
b)
F
r
equenc
y
do
ma
in m
eth
ods
S
p
a
t
i
a
l
d
o
m
a
i
n
o
p
e
r
a
t
e
s
o
n
t
h
e
p
i
x
e
l
w
h
i
l
s
t
f
r
e
q
u
e
n
c
y
d
o
m
a
i
n
a
r
e
c
o
mpu
t
ed
i
n
F
ourier
t
r
a
n
sform
in
order
to
m
o
d
i
f
y
the
freq
u
e
n
c
y
c
onte
n
t
of
t
he
i
m
a
ge
s
o
tha
t
e
d
g
e
s
an
d
o
t
he
r
su
b
tle
i
nfor
ma
t
i
o
n
can
b
e
enha
nc
e
d
[
8],
[9].
T
his
paper
sol
e
ly
f
oc
uses
o
n
spa
t
i
a
l
d
o
m
a
i
n
me
t
h
o
d
s.
T
he
t
er
m
i
t
se
l
f
w
orks
i
n
t
h
e
gi
ve
n
space
(the
ima
g
e).
This m
ea
ns
t
ha
t the
pro
c
e
dure
w
o
rks d
i
rec
tly
o
n pi
xe
ls as
show
n o
n
F
igure
3.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2502-
4752
I
n
do
n
e
si
an
J
E
l
e
c
E
n
g
&
C
o
m
p
S
ci
, V
o
l
.
1
4
,
No. 1, April 2019 :
250 –
2
57
25
2
F
i
gur
e 3.
S
pati
al dom
ai
n o
f
a
n
i
m
age
[9]
Sp
at
i
a
l
do
mai
n
e
x
p
r
es
sion
are
d
e
n
ot
ed
b
y
th
e ex
p
r
e
ssion
,
T[f(x,y)]
g(x,y)
(
1
)
Whe
r
e
g(x,y)
=
p
roce
s
s
e
d
im
a
ge
f(x,y)
= input
i
mage
T
=
oper
a
tor on
f
d
e
f
i
n
e
d
o
v
e
r so
me
n
ei
ghb
o
r
ho
od
of
(x
,y
) i
f
u
si
ng
(
x,y
)
Th
e
r
e
a
r
e
many
t
ec
hni
q
u
e
s
t
h
at
c
a
n
b
e
invo
lv
ed
i
n
spa
t
i
a
l
d
o
m
a
i
n
e
n
ha
ncem
ent
s
u
ch
a
s
co
n
t
ra
st
st
r
e
tc
hi
n
g
,
im
age
su
b
t
rac
t
i
o
n,
s
harpen
i
ng,
h
is
t
ogram
e
q
u
a
liza
t
i
o
n
,
l
o
g
tra
n
sf
orm
a
tio
ns
e
t
c
…
A
u
th
or
[
8]
c
o
ndu
c
t
ed
a
s
u
r
v
e
y
on
s
p
a
tia
l
do
mai
n
t
ech
ni
q
u
e
s
whi
c
h
i
nvo
lv
ed
p
o
int
p
r
o
c
e
s
si
ng
,
Hi
stog
ra
m
st
re
t
c
hin
g
,
H
i
sto
g
ra
m
e
q
u
a
liza
t
i
on,
s
har
p
en
i
ng
f
ilter
i
n
g
a
nd
m
a
ny
m
ore
.
B
as
ed
o
n
t
h
e
revi
ew,
t
h
e
ef
f
ecti
v
en
e
s
s
o
f
e
a
c
h
tech
n
i
q
u
es
b
e
c
o
me
s
mor
e
e
ffe
ctive
w
h
e
n
c
o
m
bine
d
m
o
re
t
han
one
m
et
ho
d.
W
h
ilst
[9]
w
o
rks
o
n
c
om
par
i
n
g
spa
tia
l
a
n
d
freque
nc
y
e
n
hanc
e
m
e
n
t
a
nd
o
b
t
a
ine
d
r
esu
lt
s
h
ow
s
t
h
a
t
spa
t
i
a
l
d
o
m
a
in
a
r
e
a
pt
f
or
s
m
a
ll
kerne
l
sinc
e F
ourier
tr
a
n
sf
orm
take
s t
i
m
e
.
O
n
the o
the
r
h
a
n
d, [1
0
] i
m
p
lem
e
nte
d
i
ma
g
e
s
ha
rpe
n
i
ng
a
nd
sm
o
o
th
in
g
b
y
fi
l
t
e
r
s
o
n
t
he
p
a
r
ti
c
u
l
a
r
im
age
.
T
he
i
n
p
u
t
im
a
g
e
s
h
ar
pe
ne
d
by
u
si
n
g
w
ei
g
h
te
d
ke
rne
l
o
f
di
ffere
n
t
v
a
l
u
e
s
a
n
d
gi
ves
s
h
ar
per
i
m
a
g
e
w
i
t
h
t
h
e
b
o
u
n
d
ary
ed
ge
i
nform
a
ti
on.
T
he
a
na
lyzed
res
u
lts
perform
ed
by
MSE
and
P
S
N
R.
R
ese
a
rc
h
b
y
[6]
surve
y
s varie
t
y o
f
con
trast
e
nha
nc
e
m
e
n
t
te
chn
i
qu
e
s
l
ik
e
Hi
st
ogra
m
e
q
u
al
i
zatio
n
(HE),
Con
t
ra
st
l
im
ite
d
A
d
a
p
tive
H
i
st
ogra
m
E
q
u
a
liz
at
ion
(
C
LA
H
E
),
M
orp
h
o
l
og
ical
e
nha
ncem
ent
o
n
a
s
in
g
l
e
sc
ale
a
n
d
Mu
lt
i
s
c
a
l
e
M
o
r
phol
o
g
i
cal
e
nh
an
ce
me
n
t
.
Co
mp
a
r
ed
r
e
s
u
l
t
o
f
t
h
ese
te
c
h
n
i
que
s
o
n
to
a
s
yn
the
tic
i
m
a
ge
o
f
ki
dne
y
a
n
d
brain
sh
ow
s
t
h
at
t
he
M
u
l
tisc
a
l
e
m
orpho
log
i
c
a
l
a
p
p
roa
ch
o
b
t
ain
e
d
resp
ect
a
b
l
e
r
esul
t
s
a
s
t
o
t
he
resul
t
s
ac
h
i
e
v
e
d
w
ith
o
t
h
er
u
l
t
r
a
m
odern
t
e
c
hn
i
que
s.
M
or
eover
,
[
1
1
]
im
pro
v
es
i
m
a
ge
qua
l
i
t
y
b
y
a
n
a
l
y
s
es
c
o
nt
ras
t
e
nh
a
n
ce
me
n
t
,
sh
a
r
pen
i
ng
a
nd
n
oi
se
r
e
d
u
c
t
i
on
o
f
t
h
e
d
a
t
ase
t
B
SD
S
3
00
B
e
rke
l
y.
Q
ua
n
titat
i
v
e
perform
ance
i
nv
o
l
ve
s
MS
E,
P
SN
R
a
nd
S
S
IM,
w
h
ich
p
r
ov
es
t
ha
t
i
m
ag
e
shar
p
e
ni
ng
pro
v
i
d
es
d
e
c
e
n
t
r
esu
lts
due
t
o
n
o
cha
n
g
e of i
nform
a
ti
on or
pi
x
el
a
s
i
t
i
s
c
l
o
s
e
t
o
t
h
e
inform
ati
o
n of
t
h
e
or
i
gi
na
l im
age
.
2.
RESEARCH
M
ETH
O
D
Th
is s
ec
ti
on
e
x
pla
i
ns
t
he
m
et
h
o
d
o
l
o
g
y
o
f t
h
i
s
paper
.
F
i
r
s
tly,
a
ppl
y in
pu
t i
m
age
o
f
M
1
t
h
e
n
,
the RG
B
i
m
ag
e
co
nv
erte
d
int
o
g
re
y
s
c
a
l
e
.
Ne
x
t
,
imp
l
ement
c
ont
ra
st
s
t
r
e
t
c
hi
n
g
o
n
the
g
i
ve
n
im
age.
T
o
fi
n
d
t
h
e
perform
ance
,
c
a
lcu
l
a
t
e
M
S
E
on
t
o
M
1
fo
llo
w
e
d
b
y
P
S
N
R.
T
he
se
s
te
p
s
are
co
nt
i
n
u
ous
u
nt
i
l
a
ll
i
n
p
u
t
i
m
a
ges
tes
t
ed
w
i
t
h
c
o
ntras
t
s
tretc
h
in
g
tec
h
n
i
que
s.
A
fte
r
t
ha
t,
i
m
p
l
e
m
e
nt
I
m
a
ge
s
ub
tra
c
t
i
on
t
e
c
h
n
i
que
s
on
to
t
he
f
o
u
r
ima
g
es
f
ol
low
e
d
b
y
I
ma
ge
s
har
p
e
n
i
ng.
F
inall
y
,
for
eac
h
t
e
ch
n
i
q
u
es,
c
a
l
c
ula
t
e
t
h
e
m
ean
v
a
l
ue
o
f
t
h
e
fou
r
ima
g
es
a
s
s
o
t
o
fin
d
w
h
i
c
h
o
f
the
thre
e
en
h
a
ncem
en
t
tec
h
ni
q
u
es
ar
e
ap
t
for
en
ha
nci
n
g
M1,
M4,
M
5
a
nd
M
7
.
F
l
ow
c
har
t
on
F
i
gure
4
de
pic
t
s cle
a
rly o
n
t
he
expla
i
ne
d m
e
t
h
o
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
ones
i
a
n
J
E
lec
En
g & Co
mp
S
c
i
IS
S
N
: 2502-
47
52
Sp
a
tia
l dom
ain
i
m
age
e
nha
nc
em
e
n
t t
e
c
h
n
i
q
u
e
s
for
ac
u
t
e
m
y
elo
i
d leu
k
e
m
ia (M1,
M4,M5,
M7) (A.S.
A
.Sal
a
m
)
25
3
.
F
i
gure
4. O
ve
ra
l
l
F
low
c
har
t
o
f
t
h
e
m
e
th
o
d
s
Inp
u
t im
ages
o
f M1,
M4,
M5
a
nd
M7
i
nser
te
d a
ccor
d
i
n
g
to
F
igure
5.
F
i
gure
5.
(
a) RG
B
M
1 (b)
RG
B
M
4
(
c
)
R
G
B
M
5
(
d
)
R
G
B
M
7
The
RG
B
ima
g
es
c
on
vert
i
n
t
o
greysca
l
e
(
F
i
gur
e
6)
t
o
reduc
e
d
i
me
n
si
on
o
f
i
m
a
g
e[12
].
A
l
s
o
,
proce
ssi
ng
b
e
c
o
m
e
s flex
i
b
le
w
hen
a
si
n
g
l
e
inte
n
s
i
t
y va
l
u
e
of
e
a
ch pixel is
specified [13].
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2502-
4752
I
n
do
n
e
si
an
J
E
l
e
c
E
n
g
&
C
o
m
p
S
ci
, V
o
l
.
1
4
,
No. 1, April 2019 :
250 –
2
57
25
4
F
i
gur
e 6.
(
a)
G
reysc
a
le
M
1
(b) Greysc
ale
M4
(
c) Gre
yscale
M5 (d
)
Gr
eyscale
M
7
Con
t
ra
st
s
tre
t
c
h
i
n
g
a
t
t
e
mpts
t
o
impro
v
e
the
con
t
ra
st
o
f
the
im
a
ge
b
y
stre
t
c
hi
n
g
r
a
n
ge
o
f
the
i
n
te
nsit
y
val
u
es.
It c
ha
n
g
es
t
he
d
is
t
r
ibu
t
i
o
n a
nd ra
nge
of d
i
g
ita
l
nu
m
b
er
s ass
i
g
n
e
d
t
o eac
h pi
xe
l
in
a
n im
age.
In m
e
dica
l
ima
g
in
g,
c
o
n
tr
ast
s
t
retc
hi
n
g
p
la
ys
a
n
i
m
p
o
r
tan
t
r
o
l
e
f
o
r
qua
l
i
t
y
e
n
hanc
e
m
e
n
t[1
4
].
F
igure
7
i
l
l
u
s
t
r
a
te
s
t
h
e
resul
t
s fr
om
s
t
r
etch
i
ng.
F
i
g
u
r
e
7
.
(
a
)
C
o
ntras
t
M
1
(b)
Co
ntras
t
M
4
(c)
Cont
r
a
st M5 (d)
C
on
t
r
ast M
7
Im
age
sub
t
ra
ct
ion
is
o
ne
o
f
t
h
e
p
o
p
u
l
a
r
m
ac
h
i
ne
v
i
s
i
on
te
chn
i
qu
e
fo
r
ex
tra
c
t
i
n
g
f
oregr
o
u
n
d
o
bj
ec
t
s
in
a
n
im
age[
15
]
.
It
i
s
o
b
t
a
i
ne
d
by
c
o
m
p
ut
ing
the
d
i
f
f
ere
n
ce
b
etw
e
e
n
a
l
l
pa
i
r
s
o
f
c
o
rres
p
o
ndi
ng
p
ix
el
s
fro
m
f
and
h
y
x
h
y
x
f
y
x
G
,
,
,
(
2
)
Whe
r
e
G
(
x,y)= I
m
a
g
e
S
ubtrac
tio
n
f(x,
y)= Im
age
ba
ck
gro
u
n
d
h(x,y)
=
I
ma
ge f
oregro
un
d
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
ones
i
a
n
J
E
lec
En
g & Co
mp
S
c
i
IS
S
N
: 2502-
47
52
Sp
a
tia
l dom
ain
i
m
age
e
nha
nc
em
e
n
t t
e
c
h
n
i
q
u
e
s
for
ac
u
t
e
m
y
elo
i
d leu
k
e
m
ia (M1,
M4,M5,
M7) (A.S.
A
.Sal
a
m
)
25
5
F
i
gur
e 8.
(a
)
Sub
trac
ti
on
M1
(
b) S
ubtrac
tio
n
M
4
(
c)
S
ubt
r
acti
o
n
M
5
(d
)
S
u
bt
ra
c
tio
n
M7
The
a
i
m
o
f
s
h
a
rpen
in
g
i
n
i
m
a
ge
p
roce
ssi
ng
i
s
t
o
e
n
h
a
n
ce
t
he
d
e
t
ai
l
s
o
f
a
n
i
ma
g
e
o
r
ev
en
l
ine
st
r
u
c
t
ures.
T
h
e
tec
h
n
i
qu
e
s
d
es
ig
ne
d
to
i
ncr
ease
t
h
e
h
i
gh
fr
eq
ue
nc
y
a
s
pe
c
t
s
o
f
t
he
i
ma
ge.
In
d
e
t
ai
ls,
sharpe
nin
g
c
o
n
ta
i
n
s
of
a
ddi
ng
s
i
gna
l
th
at
i
s
pr
op
ort
i
on
a
l
t
o
a
h
i
gh
-pass
f
ilt
e
r
ed
v
ersi
on
o
f
th
e
ori
g
i
n
al
ima
g
e.[1
1].
F
i
gur
e 8.
(
a) Sha
rpe
n
i
n
g M1
(
b)
S
harpe
n
i
n
g
M
4
(
c)
S
harpe
n
i
n
g
M5
(
d) S
har
p
en
in
g M
7
The
ima
g
es
w
ere
a
n
al
yse
d
by
us
ing
t
h
e
M
e
a
n
S
qua
re
d
Error
(
Mse
)
a
nd
P
e
ea
k
S
i
g
n
al
t
o
N
o
ise
Ra
tio(
P
S
N
R).
MS
E
repr
esent
s
t
he
c
umu
l
ati
v
e
s
quare
d
error
be
tw
ee
n
c
o
mpre
sse
d
i
m
a
ge
a
n
d
o
r
i
gi
na
l
i
m
ag
e
[
2
]
.
T
h
e
l
o
w
e
r
t
h
e
M
S
E
v
a
l
ue,
the
l
o
wer
t
h
e error r
a
te.
N
M
m,n
I
m,n
I
MSE
M,N
2
2
1
(
3
)
Whe
r
e
I
2
(m,
n)= ori
g
ina
l
im
a
g
e
I
1
(m
,
n
)
=
o
u
tp
ut
i
m
a
ge
M,
N
i
s
t
he size of
i
ma
ge
P
S
N
R
me
asure
s
t
he
p
ea
k
er
ror.
I
t
c
o
m
pute
s
t
he
p
ea
k
si
g
n
al-
t
o-n
o
ise
r
a
ti
o
i
n
d
ec
i
b
el
s
be
t
w
ee
n
tw
o
ima
g
es.
The
h
i
ghe
r t
h
e
PSNR
value,
the
b
et
te
r the
qua
li
t
y
o
f
t
h
e
im
age.
MSE
R
PSNR
2
10
log
10
(
4
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2502-
4752
I
n
do
n
e
si
an
J
E
l
e
c
E
n
g
&
C
o
m
p
S
ci
, V
o
l
.
1
4
,
No. 1, April 2019 :
250 –
2
57
25
6
Whe
r
e
R
²= 2
nᵑ-1
a
n
d
n
r
epr
e
sents n
u
m
ber
of bit
s.
I
n r
e
pre
s
e
n
t
i
ng
t
h
e
p
i
xel o
f
t
he
im
a
ge
.
3.
RESULT
S
A
N
D
ANALY
S
IS
After
t
h
e
M
S
E
a
n
d
PS
NR
h
ave
c
a
l
cu
la
ted,
t
he
r
es
u
l
t
s
a
re
t
ab
u
l
at
ed
a
s
sh
ow
n
on
T
ab
le
2
,
Ta
b
l
e
3
and
Ta
ble 4.
T
a
b
l
e 2.
Resu
lts
on C
o
n
t
r
a
st S
tre
t
c
h
i
n
g
Subt
ype
s
M
S
E
PSNR
dB
M
1
0
.
0240
6
4.
329
0
M
4
0
.
0006,
7
0.
157
9
M
5
0
.
0256
6
4.
047
4
M
7
0
.
0240
6
4
.3
2
46
Tab
l
e 3.
Re
s
u
lt
s on
Im
age S
ubtra
c
tio
n
S
ubty
p
e
s
M
S
E
PS
N
R
dB
M
1
0
.
383
1
52.
298
1
M
4
0
.
187
9
55.
391
4
M
5
0
.
516
3
51.
001
7
M
7
0
.
189
4
55.
358
0
Ta
bl
e
4
.
Re
s
ul
ts
o
n
Ima
g
e
S
h
a
r
p
e
ni
ng
Subt
ype
s
M
S
E
PS
NR
dB
M
1
0
.
0060
70.
326
1
M
4
0
.
0133
66.
900
5
M
5
0
.
0096
68.
299
3
M
7
0
.
0121
67.
307
3
S
i
nce
the
aim
of
t
h
i
s
pa
per
is
t
o
a
n
al
yz
e
an
d
se
le
ct
w
hic
h
t
ec
h
niq
u
e
s
i
s
a
pt
i
n
e
n
h
a
n
c
i
n
g
AML
su
bt
yp
e,
a
m
ea
n
v
a
lu
e
of
e
a
c
h
t
e
ch
niq
u
e
s
a
r
e
ca
l
c
u
l
at
ed
.
To
f
in
d
th
e
m
ean,
t
o
ta
l
va
lue
o
f
M
SE
a
n
d
P
S
N
R
a
r
e
di
vide
d
o
v
er
t
h
e
num
ber
of
s
ubt
y
p
es.
A
s
show
n
on eq
ua
ti
o
n
4.
4
7
5
4
1
M
M
M
M
mean
(
5
)
Tab
l
e 5.
Mea
n va
lue
s
f
or
I
m
a
ge
En
h
anc
e
m
e
nt Te
c
h
n
i
que
s
E
nha
nce
m
ent
Te
c
hnique
s
M
S
E
PS
N
R
dB
C
ont
r
a
st E
nha
n
c
em
e
n
t
0.
0186
6
5.
714
7
Im
a
g
e
Subtrac
tion
0.
3192
5
3.
512
3
Im
a
g
e
Sha
r
p
e
ning
0.
0103
6
8.
208
3
A
s
s
how
n
o
n
Table
5,
t
he
r
esults
o
f
m
e
a
n
v
a
l
ue
s
ha
ve
t
a
b
u
l
a
t
ed.
A
c
c
ordi
n
g
t
o
the
resu
lt,
Ima
g
e
sh
a
r
p
e
ni
n
g
a
c
hi
ev
ed
t
h
e
h
igh
e
st
P
S
N
R
of
6
8
.
2083
d
B.
H
e
n
ce
,
i
t
s
how
s
g
o
o
d
qua
l
i
t
y
f
or
e
nh
anc
i
n
g
A
M
L
su
b
t
y
p
es
M
1,
M
4,
M
5,
a
nd
M7.
The
te
ch
n
i
que
a
gr
ee
d
w
ith
[
1
1
]
a
s
t
h
e
au
tho
r
a
l
s
o
ac
hi
e
v
ed
i
mag
e
sharpe
nin
g
a
s a
goo
d r
e
sult
com
p
are
d
t
o ot
h
e
r
techn
i
q
u
es.
4.
CONCL
U
S
ION
Co
n
c
i
s
e
l
y
,
t
h
i
s
p
a
p
e
r
i
s
a
bout
e
xp
e
r
i
m
e
n
t
i
ng
o
n
v
a
ri
o
u
s
ima
g
e
e
nha
nc
e
m
ent
tec
hni
q
u
e
s
suc
h
a
s
con
t
ra
st
s
tre
t
c
h
in
g,
i
ma
ge
s
u
b
trac
t
i
o
n
a
n
d
i
m
a
ge
s
harpe
n
i
ng.
T
h
ese
te
ch
ni
que
s
tes
t
e
d
o
n
A
M
L
su
b
t
yp
e
M
1
,
M
4
,
M
5
a
n
d
M
7
.
E
n
h
a
n
c
i
n
g
m
e
d
i
c
a
l
i
m
a
g
e
s
i
s
e
s
s
e
n
t
i
a
l
a
s
i
t
h
e
l
p
s
to
i
m
p
ro
ve
t
he
qua
lit
y
of
t
he
i
m
a
ge
.
P
e
rform
a
nc
e
a
n
a
l
ys
is
s
uc
h
a
s
M
S
E
a
nd
P
S
N
R
h
e
l
ps
t
o
ana
l
yze
the
t
e
s
t
e
d
t
e
c
h
n
i
q
u
e
s
.
S
i
n
c
e
a
h
i
g
h
P
S
N
R
val
u
e
me
an
s
a
be
tt
e
r
qua
l
i
t
y
o
f
t
h
e
part
icu
l
ar
i
m
a
ge
,
I
m
age
s
h
a
r
p
e
n
i
n
g
show
s
g
ood
q
u
al
ity
o
u
t
o
f
the
ot
h
e
r
tw
o
tec
h
niq
u
e
s
w
ith
a
va
l
ue o
f 68.2
0
83 d
B
.
F
u
ture
w
or
k
is to te
s
t
the
pe
r
f
o
r
m
anc
e
of
t
h
e
s
e
im
ages
w
ith
ot
h
e
r
ima
g
e qua
li
ty
m
ea
sur
e
m
e
nt such
as
S
SI
M, F
1 scor
e
or e
ve
n pr
ofi
l
ing
ti
me p
erf
o
rma
n
ce
.
REFE
RENCES
[1]
M.
Y
asmin and
M.
S
hari
f
,
“Br
ain Im
ag
e En
hanc
em
en
t-A S
u
rvey
,
”
n
o
.
Janu
ary,
2
01
2.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
ones
i
a
n
J
E
lec
En
g & Co
mp
S
c
i
IS
S
N
: 2502-
47
52
Sp
a
tia
l dom
ain
i
m
age
e
nha
nc
em
e
n
t t
e
c
h
n
i
q
u
e
s
for
ac
u
t
e
m
y
elo
i
d leu
k
e
m
ia (M1,
M4,M5,
M7) (A.S.
A
.Sal
a
m
)
25
7
[2]
M
.
S
hanthi
a
nd
M
.
Ren
u
g
a
,
“
P
erf
o
rm
a
n
ce
An
a
l
ys
is
o
f
Im
age
En
han
c
em
en
t
Tec
h
ni
ques
f
o
r
ki
dney
Im
age,
”
Int
.
J.
Adv.
Res.
Elect
r
.
Elect
ron.
Ins
t
rum.
E
n
g.
,
v
o
l
.
5
,
n
o
.
5
,
p
p
.
351
7–35
22
,
2
0
1
6
.
[3
]
A.
S
.
Ab
du
l
Sala
m,
M
.
N.
M
d
.
I
sa,
M.
I
.
Ah
mad,
a
nd
R
.
Che
I
sm
ai
l,
“
Com
p
ariso
n
o
f
ed
ge
d
etect
ion
t
echn
i
q
u
es
f
or
M
7
s
ubty
p
e L
e
ukem
i
c cell in
term
s
of
no
i
s
e fi
lt
ers an
d thres
h
o
l
d v
a
l
u
e,”
EPJ W
e
b Co
nf
.
,
v
o
l
.
16
2,
20
1
7
.
[4]
a
C
arroll
et a
l
.
,
“T
he
t
(1;
2
2
)
(
p
1
3
;
q1
3)
i
s
no
n
r
an
dom
a
nd
r
est
r
ic
t
e
d
t
o
i
nfan
ts
w
ith
acu
te
m
eg
akary
obl
ast
i
c
leu
k
em
ia: a P
e
diatri
c On
co
l
o
g
y
G
roup
Stu
dy
.,”
Bl
oo
d
,
vol.
7
8
,
no.
3
,
p
p
.
74
8–7
5
2
,
1
991
.
[5
]
A.. et
.
a.
M
, “M
o
h
/
p
/ik
n
/
01
.16
(
a
r
)
,
”
Pu
t
ra
j
a
y
a
, Malay
sia.
[6]
S
.
K
.
Riti
ka,
“
C
on
trast
En
hancem
en
t
Tech
niq
u
es
f
o
r
I
m
a
ges
–
A
V
isual
Anal
y
s
is,”
Int.
J
.
Co
mput
. Appl.
,
v
o
l.
6
4,
no
.
1
7,
p
p.
2
0
–25,
2
01
3.
[7]
N.
H
a
z
w
a
ni
,
A.
H
al
im
,
M
.
Y
.
Mas
hor,
an
d
R.
H
as
san,
“
A
u
t
o
m
a
t
ic
B
la
s
t
s
C
o
un
ti
n
g
f
o
r
A
c
u
te
L
e
u
ke
mi
a
B
a
s
e
d
on
Blo
od,
”
vol.
2,
n
o.
4
,
p
p
.
9
7
1
–
976,
201
1.
[
8
]
S
.
S
.
N
e
g
i
a
n
d
B
.
G
u
p
t
a
,
“
S
u
rvey
o
f
V
a
ri
ou
s
Im
ag
e
En
han
cem
e
nt,”
Int
e
rmati
o
n
al
J. Comput.
Appl.
,
pp
.
2
2
–
3
0,
20
14
.
[9]
A. S
ara
h
, “
Im
age
En
hancem
en
t -S
p
a
tia
l
vs. F
r
eq
uency
Do
m
a
in
Fi
lters,
”
2016.
[10
]
K
.
S.
P
rav
e
en
,
K.
P
.
Bab
u
,
an
d
M
.
S
reeniv
asulu
,
“
Imp
l
e
m
en
t
ation
o
f
I
m
a
ge
S
h
a
rp
eni
ng
an
d,
”
I
n
t
.
Sci.
En
g
.
A
p
pl
.
Sci
.
,
vo
l. 2
, n
o
. 1
, pp
.
7–
1
4
,
2
01
6
.
[11
]
A
.
N
e
vriy
ant
o
a
n
d
A
.
H
.
E
q
u
a
lizati
o
n
,
“
En
hancem
en
t,
a
nd
S
t
a
n
d
a
r
d
M
e
d
i
a
n
F
i
l
t
e
r
(
N
o
i
s
e
R
e
m
o
v
a
l
)
w
i
t
h
P
i
x
e
l
-
Based
and
H
u
man
Vis
u
a
l
S
y
s
tem
-
Bas
e
d
M
eas
u
r
em
ent
s
,”
i
n
Intern
at
ion
a
l
Co
n
f
erence
on
El
ectr
i
cal E
ngineer
ing
an
d Co
mp
u
t
er
Sc
i
e
nce (
I
CE
COS
)
,
2
0
1
7
,
vol.
1
,
n
o
.
1
,
p
p
.
1
14–
119.
[1
2]
H
.
Va
gh
e
l
a
,
H
.
Mod
i
,
M.
P
a
n
dy
a
,
a
n
d
M.
P
otda
r,
“
Le
u
k
e
m
ia
D
etecti
on
u
s
i
n
g
Di
g
ital
Im
age
P
r
ocess
i
ng
Tech
niq
u
es
,”
Int
.
J. App
l
.
Inf. Sys
t
.
, vo
l
.
1
0
,
no
.
1
,
p
p
.
4
3–
51
,
20
15
.
[1
3]
K
.
Ra
gh
ul,
A.
S
.
Ra
j,
a
nd
P
.
U
.
I
la
va
ra
si,
“Ac
u
te
L
ymp
h
o
c
y
tic
L
euk
e
m
i
a
D
e
t
ecti
o
n
by
I
m
a
ge
P
ro
cessi
ng
Usin
g
Matl
a
b
,” vo
l
. 24
,
pp
.
2
63
–
2
6
7
, 20
1
6
.
[1
4]
J
.
Ka
u
r
a
n
d
A
.
Ch
ou
dh
a
r
y,
“
Co
mp
a
r
iso
n
o
f
Se
v
e
ra
l
Con
t
ra
st
St
retchi
ng
T
echn
i
ques
o
n
A
c
u
te
L
e
u
kem
i
a
Im
ag
es,
”
Int
.
J. Eng
.
In
nov
.
T
echn
o
l.
,
v
o
l
.
2,
n
o
.
1
,
pp
.
3
32–3
35,
2
0
1
2
.
[15
]
L
.
Ven
e
ts
ky,
R
.
Boczar,
an
d
R.
,
“
O
p
timi
z
atio
n
of
b
ackg
r
ou
n
d
su
bt
rac
t
io
n
f
o
r
im
age
enhan
c
em
ent,
”
Proc. S
P
I
E
Def
e
nse,
Secu
r. Sens
.
,
n
o
.
M
ay,
p
p
.
87
510
2-8
7510
2–1
3,
2
0
13.
B
I
OGRAPHIES
O
F AUTHO
RS
M
s
.
Afif
ah
S
alm
i
A
b
dul
S
al
am
i
s
a
M
a
s
t
er
b
y
Research
s
t
uden
t
i
n
U
niver
s
iti
Malaysia
P
erlis
u
n
d
e
r
the
Sch
o
o
l
o
f
M
i
cro
e
le
ctro
ni
cs.
H
e
r
f
i
el
d
o
f
i
nteres
t
i
s
M
e
d
i
cal
I
mag
e
p
ro
cess
i
ng
s
p
ecifi
call
y
i
n
Acut
e
M
y
el
o
i
d
L
e
uk
emi
a
d
e
t
ecti
on.
S
he
g
radu
ated
f
ro
m
Un
i
v
ersiti
Mal
a
ysia
Te
re
ng
ga
n
u
,
in
2
01
6
with
a
B
a
c
he
lor
De
g
r
ee
of
A
pp
lie
d
S
c
i
e
n
c
e
(
P
h
ys
ics,
E
l
ectro
n
i
c
and
Instrum
e
nt
at
ion)
.
D
r
.
M
o
hd
N
az
rin
M
d
I
sa
i
s
a
seni
or
l
ectu
r
e
r
i
n
the
S
c
ho
ol
o
f
M
i
cro
e
lect
roni
c
Eng
i
n
eering
at
U
n
iv
ersi
ti
M
alaysia
P
e
rlis
(
UniM
A
P
).
C
urren
tly,
he
i
s
a
mem
b
er
of
I
n
t
eg
rated
Circui
ts
a
nd
S
y
ste
m
s
D
e
s
i
gn
(ICAS
e)
g
rou
p
.
His
research
i
nteres
ts
i
nclu
de
r
e
con
f
igu
r
abl
e
a
rch
itect
ures
,
b
i
o
i
n
f
o
r
m
a
ti
c
s
a
nd
c
om
p
u
tatio
n
a
l
bio
l
o
gy,
f
iel
d
p
rogram
m
a
b
l
e
ga
t
e
a
rray
(FPGA
)
a
nd
ASI
C
d
e
sig
n
.
H
e
grad
uated
his
do
c
t
orat
e
s
t
u
d
y
f
r
om
t
he
U
n
i
v
e
rs
it
y
o
f
E
din
b
u
r
gh,
S
c
o
tl
and,
UK
in
2
0
1
3
.
H
i
s
PhD th
e
s
is
e
n
t
itl
ed
"
H
i
gh
P
erf
o
rm
a
n
ce
Re
co
nf
ig
urab
le
Archi
t
ect
ures
f
or
B
io
logica
l
S
e
q
u
ence A
l
i
gnm
e
n
ts
"
D
r
M
uh
am
m
a
d
Imran
A
h
m
a
d
rec
e
iv
ed
h
is
P
h
D
i
n
Com
p
u
t
er
E
ngi
neeri
n
g
from
New
c
a
s
tle
Univers
i
ty,
United King
dom
i
n
2
014. Current
ly
h
e is a
s
enior
le
cturer at
S
c
ho
ol
o
f Co
m
put
er and
Commun
i
cat
ion
E
n
gineeri
n
g,
U
niversiti
Mal
a
ysi
a
P
erl
i
s.
H
i
s
r
ese
arch
i
nt
erests
i
n
c
lud
e
b
i
o
m
e
t
ric,
s
i
gnal
anal
ys
is
a
nd
imag
e
p
ro
cessi
ng
.
Evaluation Warning : The document was created with Spire.PDF for Python.