I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
12
,
No
.
1
,
Feb
r
u
ar
y
20
22
,
p
p
.
585
~
595
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/
ijece
.
v
12
i
1
.
pp
5
8
5
-
5
9
5
585
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
M
o
delling
on
-
de
ma
nd preproces
si
ng
f
ra
mewo
rk
towa
rds
pra
ctical a
ppro
a
c
h in c
lin
ica
l ana
ly
sis
of diabetic
re
ti
no
pa
thy
P
ra
k
rut
hi M
a
nd
y
a
K
rish
neg
o
wda
1
,
K
o
ma
ra
s
a
m
y
G
a
nes
a
n
2
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
Ja
i
n
(
D
e
e
m
e
d
t
o
b
e
U
n
i
v
e
r
s
i
t
y
)
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
2
S
c
h
o
o
l
o
f
C
o
m
p
u
t
i
n
g
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
r
i
n
g
,
V
I
T
B
h
o
p
a
l
U
n
i
v
e
r
si
t
y
,
S
e
h
o
r
e
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
an
1
,
2
0
2
1
R
ev
is
ed
J
u
l 1
5
,
2
0
2
1
Acc
ep
ted
Au
g
4
,
2
0
2
1
Dia
b
e
ti
c
re
ti
n
o
p
a
t
h
y
(DR)
re
fe
rs
to
a
c
o
m
p
li
c
a
ti
o
n
o
f
d
iab
e
tes
a
n
d
a
p
rime
c
a
u
se
o
f
v
isio
n
l
o
ss
in
m
id
d
le
-
a
g
e
d
p
e
o
p
le.
A t
ime
ly
sc
re
e
n
in
g
a
n
d
d
iag
n
o
sis
p
ro
c
e
ss
c
a
n
re
d
u
c
e
th
e
risk
o
f
b
l
i
n
d
n
e
ss
.
F
u
n
d
u
s
ima
g
i
n
g
is
m
a
in
l
y
p
re
fe
rre
d
in
t
h
e
c
li
n
ica
l
a
n
a
ly
sis
o
f
DR
.
H
o
we
v
e
r;
t
h
e
ra
w
fu
n
d
u
s
ima
g
e
s
a
re
u
su
a
ll
y
su
b
jec
ted
to
a
rti
fa
c
ts,
n
o
ise
,
l
o
w
a
n
d
v
a
ried
c
o
n
tras
t,
w
h
ich
is
v
e
r
y
h
a
rd
to
p
ro
c
e
ss
b
y
h
u
m
a
n
v
is
u
a
l
sy
ste
m
s
a
n
d
a
u
t
o
m
a
ted
sy
ste
m
s.
In
th
e
e
x
isti
n
g
li
tera
tu
re
,
m
a
n
y
so
l
u
ti
o
n
s
a
re
g
iv
e
n
to
e
n
h
a
n
c
e
th
e
f
u
n
d
u
s
ima
g
e
.
Ho
we
v
e
r,
su
c
h
a
p
p
r
o
a
c
h
e
s
a
re
p
a
rti
c
u
lar
a
n
d
li
m
it
e
d
t
o
a
sp
e
c
ifi
c
o
b
jec
ti
v
e
t
h
a
t
c
a
n
n
o
t
a
d
d
re
ss
m
u
lt
i
p
le
fu
n
d
u
s
ima
g
e
s.
Th
is
p
a
p
e
r
h
a
s
p
re
se
n
ted
a
n
o
n
-
d
e
m
a
n
d
p
re
p
ro
c
e
ss
in
g
fra
m
e
wo
rk
t
h
a
t
in
teg
ra
tes
d
iffere
n
t
tec
h
n
iq
u
e
s
to
a
d
d
re
ss
g
e
o
m
e
tri
c
a
l
issu
e
s,
ra
n
d
o
m
n
o
ise
s,
a
n
d
c
o
m
p
r
e
h
e
n
si
v
e
c
o
n
tras
t
e
n
h
a
n
c
e
m
e
n
t
so
lu
ti
o
n
s.
T
h
e
p
e
rfo
rm
a
n
c
e
o
f
e
a
c
h
p
re
p
ro
c
e
ss
in
g
p
r
o
c
e
ss
is
e
v
a
lu
a
ted
a
g
a
in
st
p
e
a
k
sig
n
a
l
-
to
-
n
o
ise
ra
ti
o
(
P
S
NR
)
,
a
n
d
b
r
ig
h
tn
e
ss
is
q
u
a
n
t
if
ied
in
th
e
e
n
h
a
n
c
e
d
ima
g
e
.
T
h
e
m
o
ti
v
e
o
f
t
h
is
p
a
p
e
r
is
t
o
o
ffe
r
a
flex
i
b
le
a
p
p
ro
a
c
h
o
f
p
re
p
ro
c
e
ss
in
g
m
e
c
h
a
n
ism
t
h
a
t
c
a
n
m
e
e
t
ima
g
e
e
n
h
a
n
c
e
m
e
n
t
n
e
e
d
s
b
a
se
d
o
n
d
iffere
n
t
p
re
p
ro
c
e
ss
in
g
re
q
u
ire
m
e
n
ts
to
imp
ro
v
e
th
e
q
u
a
li
t
y
o
f
fu
n
d
u
s
ima
g
in
g
to
wa
rd
s e
a
rly
-
sta
g
e
d
iab
e
ti
c
re
ti
n
o
p
a
th
y
id
e
n
ti
fica
ti
o
n
.
K
ey
w
o
r
d
s
:
C
L
AHE
Diab
etic
r
etin
o
p
ath
y
PS
NR
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Pra
k
r
u
th
i M
an
d
y
a
Kr
is
h
n
e
g
o
wd
a
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
J
ain
(
Dee
m
ed
to
b
e
Un
iv
er
s
ity
)
B
en
g
alu
r
u
,
I
n
d
ia
E
m
ail:
m
k
p
r
ak
r
u
th
i.1
0
1
2
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
Diab
etic
R
etin
o
p
ath
y
r
ef
er
s
to
a
v
is
io
n
d
is
o
r
d
er
ca
u
s
ed
b
y
a
m
etab
o
lic
d
is
ea
s
e
ca
lled
d
iab
etes,
wh
ich
ca
u
s
es
v
is
io
n
d
am
ag
e
r
an
g
in
g
f
r
o
m
b
lu
r
r
ed
v
is
io
n
to
co
m
p
lete
b
lin
d
n
ess
in
d
iab
etic
p
atien
ts
[
1
]
,
[
2
]
.
Diab
etic
r
etin
o
p
ath
y
(
DR
)
p
r
o
g
r
ess
es
in
th
r
ee
s
tag
es
a
m
in
o
r
,
m
o
d
er
ate,
an
d
ad
v
an
ce
d
s
tag
e,
ca
u
s
in
g
r
etin
al
is
ch
em
ia
an
d
co
m
p
lete
b
lin
d
n
ess
,
wh
ich
is
ch
ar
ac
ter
iz
ed
b
y
th
e
ap
p
ea
r
an
ce
o
f
s
y
m
p
to
m
s
s
u
c
h
as
m
icr
o
an
eu
r
y
s
m
s
an
d
e
x
u
d
ate
s
an
d
m
an
y
o
th
e
r
p
ath
o
lo
g
ica
l
f
ea
tu
r
es
[
3
]
.
E
f
f
ec
tiv
e
s
cr
ee
n
in
g
an
d
d
ia
g
n
o
s
is
ar
e
r
eq
u
ir
e
d
in
o
p
h
t
h
alm
o
lo
g
y
clin
ical
an
aly
s
is
to
d
etec
t
D
R
an
d
s
u
itab
le
tr
ea
tm
en
t
at
an
ea
r
ly
s
tag
e.
T
o
d
ate,
lo
ts
o
f
m
ed
ical
im
ag
in
g
tech
n
iq
u
es
ar
e
a
v
ailab
le
in
th
e
b
io
m
ed
ical
f
ield
,
a
m
o
n
g
wh
ich
f
u
n
d
u
s
im
a
g
in
g
is
th
e
b
est
av
ailab
le
im
ag
in
g
m
o
d
al
ity
in
o
p
h
th
alm
o
l
o
g
y
to
ca
r
r
y
o
u
t
d
iag
n
o
s
is
an
d
id
en
tific
ati
o
n
o
f
DR
[
4
]
,
[
5
]
.
Ho
wev
er
,
o
n
e
o
f
th
e
m
aj
o
r
is
s
u
es
en
co
u
n
ter
ed
in
d
etec
tin
g
b
lo
o
d
v
ess
els
an
d
d
iag
n
o
s
in
g
p
ath
o
lo
g
ical
s
ig
n
s
is
th
e
p
o
o
r
im
a
g
e
q
u
ality
d
u
e
t
o
p
o
o
r
co
n
tr
ast,
u
n
ev
e
n
illu
m
in
atio
n
,
a
n
d
n
o
is
e
in
clu
s
io
n
d
u
r
in
g
th
e
f
u
n
d
u
s
im
ag
e
ac
q
u
is
itio
n
p
r
o
ce
s
s
[
6
]
.
T
h
er
ef
o
r
e
,
a
p
r
ep
r
o
ce
s
s
in
g
m
ec
h
an
is
m
is
n
ee
d
ed
f
o
r
ex
ec
u
tin
g
ef
f
ec
tiv
e
im
ag
e
s
eg
m
en
tatio
n
an
d
an
ac
cu
r
ate
d
is
ea
s
e
id
en
tific
atio
n
p
r
o
c
ess
.
As
a
p
r
im
ar
y
s
tep
in
p
r
ep
r
o
ce
s
s
in
g
,
im
ag
e
en
h
an
ce
m
e
n
t
is
ca
r
r
ied
o
u
t
to
im
p
r
o
v
e
th
e
co
n
tr
ast
a
n
d
v
iv
id
n
ess
o
f
th
e
f
ea
tu
r
es
in
th
e
f
u
n
d
u
s
im
a
g
es
[
7
]
.
C
h
an
g
es
in
th
e
elem
en
ts
an
d
tex
tu
r
es
o
f
o
b
jects
in
th
e
en
h
an
ce
d
f
u
n
d
u
s
im
ag
es
p
r
o
v
id
e
an
ess
en
tial
b
io
m
ar
k
e
r
f
o
r
ca
r
r
y
i
n
g
o
u
t
c
o
m
p
r
eh
e
n
s
iv
e
clin
ical
a
n
aly
s
is
o
f
th
e
r
eg
i
o
n
o
f
in
ter
est
(
R
OI
)
to
war
d
s
p
r
e
-
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
12
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
5
8
5
-
595
586
ev
alu
atio
n
o
f
p
atien
t
an
d
d
ia
g
n
o
s
is
o
f
DR
[
8
]
.
T
h
er
e
is
a
r
an
g
e
o
f
p
r
e
p
r
o
ce
s
s
in
g
ap
p
r
o
ac
h
es
g
iv
en
in
th
e
liter
atu
r
e
f
o
r
e
n
h
an
cin
g
t
h
e
q
u
ality
o
f
th
e
f
u
n
d
u
s
im
a
g
es.
Ga
m
m
a
co
r
r
ec
tio
n
an
d
h
is
to
g
r
a
m
-
b
ased
tech
n
iq
u
es
ar
e
ex
ten
s
iv
ely
ad
o
p
ted
to
im
p
r
o
v
e
f
u
n
d
u
s
im
ag
es'
v
is
u
al
c
h
ar
ac
ter
is
tics
d
u
e
to
th
eir
ad
m
ir
ab
le
p
er
f
o
r
m
an
ce
[9
]
,
[
1
0
]
.
Alth
o
u
g
h
th
e
r
e
is
v
ar
io
u
s
av
ailab
ilit
y
o
f
p
r
e
p
r
o
c
ess
in
g
m
ed
ical
o
b
jects,
th
er
e
is
less
ev
id
en
ce
o
f
p
er
f
o
r
m
in
g
ef
f
ec
tiv
e
p
r
e
p
r
o
c
ess
in
g
to
war
d
s
th
e
r
etin
a.
T
h
er
ef
o
r
e
,
th
is
p
ap
er
p
r
esen
ts
a
d
is
cu
s
s
io
n
o
f
a
co
m
p
u
tatio
n
al
m
o
d
el
o
f
p
r
e
p
r
o
ce
s
s
in
g
aim
in
g
to
war
d
s
D
R
.
T
h
e
m
an
u
s
cr
ip
t'
s
o
r
g
an
iza
tio
n
is
as
f
o
l
lo
ws:
Sectio
n
2
d
is
cu
s
s
es
th
e
ex
is
tin
g
s
tu
d
ies
in
th
e
f
o
r
m
o
f
th
e
s
tu
d
y
b
ac
k
g
r
o
u
n
d
,
s
ec
tio
n
3
h
ig
h
lig
h
ts
th
e
r
esear
ch
p
r
o
b
lem
s
,
wh
ile
Sectio
n
4
b
r
i
ef
s
ab
o
u
t
p
r
o
p
o
s
ed
m
eth
o
d
o
l
o
g
y
in
ter
m
s
o
f
s
o
lu
tio
n
.
Sectio
n
5
d
is
cu
s
s
es
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
wh
ile
s
ec
tio
n
6
d
is
cu
s
s
es th
e
r
esu
lt a
n
d
d
is
cu
s
s
io
n
,
an
d
s
ec
tio
n
7
d
is
cu
s
s
es th
e
co
n
clu
s
io
n
o
f
th
e
s
tu
d
y
.
2.
RE
L
AT
E
D
WO
RK
T
h
is
s
ec
tio
n
d
is
cu
s
s
e
s
ex
is
t
in
g
s
tu
d
ies
to
war
d
s
p
r
ep
r
o
ce
s
s
in
g
.
Xiao
et
a
l.
[
1
1
]
ad
o
p
ted
a
g
am
m
a
co
r
r
ec
tio
n
m
ec
h
an
is
m
f
o
r
th
e
f
u
n
d
u
s
im
ag
e
en
h
an
ce
m
e
n
t
f
o
r
th
e
au
to
m
atic
d
etec
tio
n
o
f
DR
ex
tr
ac
tin
g
th
e
g
r
ee
n
co
m
p
o
n
en
t o
f
th
e
f
u
n
d
u
s
im
ag
e
an
d
d
eter
m
in
ed
th
e
h
i
s
to
g
r
am
cu
r
v
e
s
lo
p
e.
Mo
u
za
i
et
a
l.
[
1
2
]
p
r
esen
ted
a
f
u
zz
y
-
b
ased
g
am
m
a
co
r
r
ec
tio
n
m
ec
h
a
n
is
m
f
o
r
b
r
ig
h
tn
ess
p
r
eser
v
atio
n
.
Acc
o
r
d
in
g
to
th
e
s
tu
d
y
o
f
[
1
3
]
,
it
is
f
o
u
n
d
th
at
h
is
to
g
r
am
-
b
ased
te
ch
n
iq
u
es,
esp
ec
ially
h
is
to
g
r
a
m
eq
u
aliza
tio
n
(
HE
)
a
n
d
c
o
n
t
r
ast
lim
ited
ad
ap
tiv
e
h
is
to
g
r
am
eq
u
aliza
tio
n
,
ar
e
m
ain
ly
u
s
ed
to
in
cr
ea
s
e
co
n
tr
ast
m
ed
ical
im
ag
es.
Ho
wev
er
,
c
o
n
t
r
a
s
t
l
i
m
i
t
e
d
a
d
a
p
t
i
v
e
h
i
s
t
o
g
r
a
m
e
q
u
a
l
i
z
a
t
i
o
n
(
C
L
A
H
E
)
i
s
i
n
t
r
o
d
u
c
e
d
a
s
a
n
i
m
p
r
o
v
i
s
e
d
v
e
r
s
i
o
n
o
f
H
E
t
h
a
t
u
s
e
s
a
c
l
i
p
l
i
m
i
t
i
n
g
m
e
c
h
a
n
i
s
m
,
w
h
i
c
h
r
e
d
u
c
e
s
t
h
e
o
v
e
r
b
r
i
g
h
t
n
e
s
s
a
n
d
p
r
o
v
i
d
e
s
b
e
t
t
e
r
e
n
h
a
n
c
e
m
e
n
t
r
e
s
u
l
t
s
[
1
4
]
.
Yad
av
et
a
l.
[
1
5
]
m
en
tio
n
ed
HE
is
s
u
b
jecte
d
to
am
p
lific
atio
n
o
f
n
o
is
e
an
d
o
v
er
en
h
an
ce
m
en
t
f
ac
to
r
.
Sh
a
m
s
u
d
ee
n
et
a
l.
[
1
6
]
in
tr
o
d
u
ce
d
an
im
p
r
o
v
e
d
HE
m
ec
h
an
is
m
to
ad
ju
s
t
b
r
i
g
h
tn
e
s
s
an
d
n
o
is
e
elim
in
atio
n
.
T
h
e
ad
o
p
tio
n
o
f
a
j
o
in
t
ap
p
r
o
ac
h
o
f
g
am
m
a
c
o
r
r
ec
tio
n
an
d
C
L
AHE
is
f
o
u
n
d
in
th
e
s
tu
d
y
o
f
Z
h
o
u
et
a
l.
[
1
7
]
.
I
n
th
is
wo
r
k
,
th
e
lu
m
in
an
ce
g
ain
m
atr
ix
is
co
n
s
tr
u
cted
b
y
a
p
p
ly
in
g
a
g
am
m
a
co
r
r
ec
tio
n
m
ec
h
an
is
m
in
th
e
HSV
co
lo
r
s
p
ac
e
f
o
r
ea
ch
co
lo
r
c
o
m
p
o
n
en
t
o
f
th
e
i
n
p
u
t im
ag
e
.
Fu
r
th
er
,
th
e
s
ec
o
n
d
ar
y
lev
el
o
f
co
n
tr
ast
ad
ju
s
tm
en
t
is
ca
r
r
ie
d
o
u
t
in
t
h
e
ch
an
n
el,
wh
er
e
C
L
AHE
is
ap
p
lied
o
v
er
t
h
e
L
AB
co
lo
r
s
p
ac
e.
T
h
is
tech
n
iq
u
e
h
as
d
em
o
n
s
tr
ated
a
n
in
tellig
en
t
a
p
p
r
o
ac
h
o
f
g
a
m
m
a
co
r
r
ec
tio
n
an
d
C
L
AHE
to
g
et
h
er
to
en
h
a
n
ce
th
e
co
n
tr
ast
o
f
th
e
f
u
n
d
u
s
im
a
g
e.
A
p
ar
t
f
r
o
m
u
s
i
n
g
c
o
n
t
r
a
s
t
a
d
j
u
s
t
m
e
n
t
t
e
c
h
n
i
q
u
e
s
,
f
i
l
t
e
r
i
n
g
m
e
c
h
a
n
i
s
m
s
s
u
c
h
a
s
m
e
a
n
a
n
d
m
e
d
i
a
n
f
i
l
t
e
r
i
n
g
a
r
e
c
o
n
s
i
d
e
r
e
d
t
o
e
l
i
m
i
n
a
t
e
t
h
e
f
u
n
d
u
s
i
m
a
g
e
n
o
i
s
e
s
.
A
c
o
m
b
i
n
a
t
i
o
n
o
f
n
o
i
s
e
f
i
l
t
e
r
s
a
n
d
C
L
A
H
E
i
s
c
o
n
s
i
d
e
r
e
d
i
n
t
h
e
s
t
u
d
y
o
f
Ku
m
ar
et
a
l.
[
1
8
]
.
T
h
is
s
tu
d
y
u
s
es
a
f
u
zz
y
-
b
ased
m
ed
ian
f
ilter
an
d
b
lo
c
k
m
atch
in
g
alg
o
r
ith
m
to
p
e
r
f
o
r
m
n
o
is
e
elim
in
atio
n
.
C
L
AHE
an
d
th
e
m
o
d
if
ied
HE
tech
n
iq
u
e
en
h
a
n
ce
th
e
co
n
tr
a
s
t
o
f
th
e
f
ilter
ed
im
ag
e.
I
n
a
s
im
ilar
d
ir
ec
tio
n
,
th
e
wo
r
k
d
o
n
e
b
y
So
n
ali
et
a
l.
[
1
9
]
u
s
ed
a
c
o
m
b
in
ati
o
n
o
f
v
ar
i
o
u
s
n
o
is
e
f
ilter
s
s
u
ch
as
m
ea
n
,
m
ed
ian
,
Gau
s
s
ian
,
an
d
C
L
AHE
ar
e
th
en
ap
p
lied
to
th
e
d
en
o
is
ed
im
ag
e
c
o
lo
r
c
o
m
p
o
n
en
ts
.
Ad
o
p
tin
g
th
e
h
az
e
r
em
o
v
al
alg
o
r
ith
m
is
also
s
ee
n
in
Vin
o
d
h
i
n
i
et
a
l
.
[
2
0
]
t
o
ad
j
u
s
t
u
n
-
ev
en
illu
m
i
n
atio
n
in
f
u
n
d
u
s
im
ag
e
p
ix
el
i
n
ten
s
ity
.
T
h
e
ed
g
e
is
s
m
o
o
th
ed
b
y
u
s
in
g
a
f
ilter
in
g
m
ec
h
an
is
m
.
R
ed
d
y
et
a
l.
[
2
1
]
p
r
esen
ted
an
in
te
g
r
ated
en
h
a
n
ce
m
en
t a
lg
o
r
ith
m
to
p
er
f
o
r
m
o
p
tim
al
en
h
an
ce
m
en
t
o
f
r
etin
al
im
ag
es.
T
h
e
au
th
o
r
s
h
av
e
u
s
ed
a
f
lig
h
t
f
ir
ef
ly
o
p
tim
izatio
n
ap
p
r
o
ac
h
with
ad
ap
tiv
e
g
am
m
a
-
co
r
r
ec
ted
a
n
d
tex
tu
r
e
HE
f
r
am
ewo
r
k
f
o
r
im
p
r
o
v
in
g
o
v
er
all
v
is
u
al
q
u
ality
.
Ho
wev
er
,
th
is
s
tu
d
y
witn
ess
es
co
m
p
u
tatio
n
al
c
o
m
p
lex
ity
as
it
p
er
f
o
r
m
s
a
m
u
lti
-
lev
el
co
m
p
u
tatio
n
f
o
r
d
if
f
er
en
t
tech
n
iq
u
es.
Fra
z
et
a
l.
[
2
2
]
p
r
esen
ted
a
s
u
r
v
ey
o
f
e
x
is
tin
g
m
eth
o
d
s
f
o
r
e
x
tr
ac
tin
g
th
e
r
etin
al
v
ess
el.
Mo
h
an
ty
et
a
l.
[
2
3
]
an
d
Dai
et
a
l.
[
2
4
]
ca
r
r
ie
d
e
x
ten
s
iv
e
in
v
esti
g
atio
n
an
d
an
al
y
s
is
o
f
th
e
f
u
n
d
u
s
im
ag
e
s
cr
ee
n
in
g
an
d
au
to
m
atic
DR
id
en
tific
atio
n
d
iag
n
o
s
is
.
Kh
an
et
a
l.
[
2
5
]
p
r
esen
ted
a
r
ev
iew
wo
r
k
o
n
th
e
f
u
n
d
u
s
im
ag
in
g
tec
h
n
iq
u
es
to
war
d
s
th
e
p
r
ec
is
e
e
x
tr
ac
tio
n
o
f
b
l
o
o
d
v
ess
els.
L
in
et
a
l.
[
2
6
]
c
ar
r
ied
a
s
y
s
tem
atic
an
aly
s
is
o
f
r
etin
al
im
ag
es
f
o
r
t
h
e
au
to
m
atic
s
cr
ee
n
in
g
o
f
th
e
DR
.
3.
P
RO
B
L
E
M
ST
A
T
E
M
E
NT
T
h
e
r
esear
ch
p
r
o
b
lem
s
b
ased
o
n
th
e
a
b
o
v
e
-
r
elate
d
wo
r
k
ar
e
an
aly
ze
d
as f
o
llo
ws:
−
I
r
r
esp
ec
tiv
e
o
f
ex
te
n
s
iv
e
r
esear
ch
to
war
d
s
DR
u
s
in
g
f
u
n
d
u
s
im
ag
es
wh
ile
f
o
cu
s
in
g
o
n
ad
d
r
ess
in
g
en
h
an
ce
m
e
n
t iss
u
es,
th
er
e
ar
e
f
ewe
r
s
tan
d
ar
d
m
o
d
els.
−
Usag
e
o
f
im
a
g
e
en
h
an
ce
m
e
n
t
tech
n
iq
u
es
is
eith
er
ca
r
r
ied
in
th
e
lim
ited
s
co
p
e
o
f
p
r
e
p
r
o
ce
s
s
in
g
o
p
er
atio
n
.
I
n
ad
d
itio
n
,
s
o
m
e
ar
e
ass
o
ciate
d
with
a
co
m
p
u
tatio
n
al
c
o
m
p
lex
ity
th
at
d
o
esn
'
t
m
ee
t
th
e
r
eq
u
ir
e
m
en
t
o
f
ef
f
ec
tiv
e
d
iag
n
o
s
is
an
d
DR
s
cr
ee
n
in
g
.
−
T
h
e
ex
is
tin
g
en
h
an
ce
m
en
t
m
e
ch
an
is
m
is
ca
r
r
ied
o
u
t
h
ig
h
l
y
r
ec
u
r
s
iv
e,
wh
ich
a
s
im
p
le
an
d
tim
e
-
ef
f
icien
t
im
p
lem
en
tatio
n
d
esig
n
n
ee
d
s
ap
p
r
o
ac
h
f
o
r
f
u
n
d
u
s
im
ag
es.
−
Als
o
,
th
er
e
is
le
s
s
av
ailab
ilit
y
o
f
jo
in
t
m
ec
h
an
is
m
s
t
o
ad
d
r
e
s
s
s
ev
er
al
en
h
an
ce
m
en
t
p
r
o
b
l
em
s
in
a
s
in
g
le
m
o
d
u
le.
Actu
ally
,
th
e
en
h
a
n
c
em
en
t
is
ca
r
r
ied
o
u
t
to
s
p
ec
if
ic
o
b
jectiv
es,
m
ain
ly
to
r
esear
ch
wo
r
k
.
T
h
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mo
d
ellin
g
o
n
d
ema
n
d
p
r
ep
r
o
c
ess
in
g
fr
a
mewo
r
k
to
w
a
r
d
s
…
(
P
r
a
kru
th
i Ma
n
d
ya
K
r
is
h
n
eg
o
w
d
a
)
587
liter
atu
r
e
lack
s
a
co
m
p
r
eh
e
n
s
iv
e
s
et
o
f
p
r
ep
r
o
ce
s
s
in
g
ap
p
r
o
ac
h
es
to
m
ee
t
all
im
ag
e
en
h
an
ce
m
e
n
t
r
eq
u
ir
em
e
n
ts
in
a
s
in
g
le
d
ep
l
o
y
m
en
t scen
ar
io
.
T
h
er
ef
o
r
e,
th
e
p
r
o
b
lem
s
tatem
en
t
f
o
r
th
e
p
r
o
p
o
s
ed
s
y
s
tem
ca
n
b
e
s
tated
as
"De
v
elo
p
in
g
a
co
m
p
u
tatio
n
al
m
o
d
el
th
at
ca
n
p
r
o
v
id
e
o
n
-
d
em
an
d
p
r
ep
r
o
ce
s
s
in
g
ap
p
r
o
ac
h
es
to
ad
d
r
ess
m
u
ltip
le
en
h
a
n
ce
m
en
t
p
r
o
b
lem
s
ass
o
ciate
d
with
f
u
n
d
u
s
im
ag
e
is
a
ch
allen
g
in
g
task
.
"
4.
P
RO
P
O
SE
D
SO
L
UT
I
O
N
T
h
e
p
r
o
p
o
s
ed
r
esear
ch
s
tu
d
y
in
ten
d
ed
to
p
r
esen
t
a
u
n
iq
u
e
s
o
lu
tio
n
to
ad
d
r
ess
d
if
f
er
en
t
p
r
o
b
lem
s
r
elate
d
to
im
ag
e
q
u
ality
with
i
n
a
s
in
g
le
s
et
o
f
c
o
m
p
u
tatio
n
a
l
m
o
d
els.
T
h
e
c
u
r
r
e
n
t
r
esear
c
h
wo
r
k
p
r
esen
ts
th
e
co
m
p
u
tatio
n
al
f
r
am
ewo
r
k
m
o
d
elin
g
,
wh
ich
o
f
f
er
s
on
-
d
em
an
d
p
r
e
p
r
o
ce
s
s
in
g
s
er
v
ices
to
ca
r
r
y
o
u
t
co
m
p
r
eh
e
n
s
iv
e
en
h
an
ce
m
en
t
o
p
er
atio
n
s
o
v
er
f
u
n
d
u
s
im
a
g
es.
T
h
e
o
n
-
d
e
m
an
d
p
r
ep
r
o
c
ess
in
g
is
m
ea
n
t
to
f
ac
ilit
ate
a
v
ar
iety
o
f
im
ag
e
q
u
ality
im
p
r
o
v
em
e
n
t
m
ec
h
a
n
is
m
s
in
a
s
in
g
le
ap
p
licatio
n
,
with
an
ad
v
a
n
tag
e
o
v
er
m
u
ltip
le
p
r
ep
r
o
ce
s
s
in
g
r
eq
u
ir
em
en
ts
as
p
er
th
e
v
is
u
al
ch
ar
ac
ter
is
tics
o
f
th
e
f
u
n
d
u
s
im
ag
in
g
.
T
h
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
in
teg
r
ates
d
if
f
er
e
n
t
p
r
ep
r
o
ce
s
s
in
g
alg
o
r
ith
m
s
.
T
h
e
s
elec
tio
n
o
f
s
u
itab
le
en
h
an
ce
m
en
t
as
a
p
r
ac
tical
s
o
lu
tio
n
is
b
ased
o
n
e
v
alu
atin
g
ea
ch
p
r
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
e
co
n
ce
r
n
in
g
a
p
ar
ticu
lar
in
p
u
t
f
u
n
d
u
s
im
a
g
e.
T
h
e
p
er
f
o
r
m
an
ce
ev
alu
atio
n
is
ca
r
r
ied
b
ased
o
n
th
e
en
h
a
n
ce
d
i
m
ag
e
s
tatis
tic
s
an
d
q
u
ality
m
etr
ics
s
u
ch
as
p
ea
k
s
ig
n
al
-
to
-
n
o
is
e
r
atio
(
PS
NR
)
an
d
B
r
ig
h
tn
ess
as
a
p
e
r
f
o
r
m
a
n
ce
in
d
icato
r
f
o
r
ea
ch
p
r
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
e.
T
h
e
ar
ch
itectu
r
e
o
f
th
e
o
n
-
d
e
m
an
d
p
r
ep
r
o
ce
s
s
in
g
f
r
am
ewo
r
k
is
d
ep
icted
in
Fig
u
r
e
1
.
I
n
p
u
t
F
u
n
d
u
s
I
m
a
g
e
G
e
o
m
e
t
r
i
c
a
l
A
d
j
u
s
t
m
e
n
t
F
i
l
t
e
r
i
n
g
C
o
n
t
r
a
s
t
E
n
h
a
n
c
e
m
e
n
t
S
c
a
l
i
n
g
,
C
r
o
p
p
i
n
g
,
N
o
r
m
a
l
i
z
a
t
i
o
n
M
e
a
n
F
i
l
t
e
r
,
M
e
d
i
a
n
F
i
l
t
e
r
P
o
w
e
r
l
a
w
,
C
L
A
H
E
,
C
o
l
o
r
E
n
h
a
n
c
e
m
e
n
t
,
M
a
t
h
e
m
a
t
i
c
a
l
M
o
r
p
h
o
l
o
g
y
,
O
p
t
i
m
i
z
a
t
i
o
n
E
n
h
a
n
c
e
d
F
u
n
d
u
s
I
m
a
g
e
P
e
r
f
o
r
m
a
n
c
e
m
e
t
r
i
c
s
:
P
S
N
R
,
B
r
i
g
h
t
n
e
s
s
Fig
u
r
e
1
.
Ou
tlin
e
o
f
th
e
p
r
o
p
o
s
ed
p
r
e
-
p
r
o
ce
s
s
in
g
tech
n
iq
u
es a
n
d
ev
alu
atio
n
m
o
d
el
Fig
u
r
e
1
d
e
p
icts
th
e
ar
c
h
itectu
r
e
o
f
th
e
p
r
o
p
o
s
ed
o
n
-
d
e
m
an
d
p
r
ep
r
o
ce
s
s
in
g
f
r
am
ew
o
r
k
,
wh
ich
in
co
r
p
o
r
ates
d
if
f
er
en
t
p
r
ep
r
o
c
ess
in
g
tech
n
iq
u
es
s
u
ch
as
in
ter
p
o
latio
n
,
cr
o
p
p
i
n
g
,
an
d
n
o
r
m
alizin
g
to
ad
d
r
ess
g
eo
m
etr
ical
d
ef
o
r
m
atio
n
s
.
I
n
ad
d
itio
n
,
m
ea
n
f
ilter
in
g
,
m
ed
ian
f
ilter
in
g
,
an
d
r
ec
u
r
s
iv
e
f
ilter
in
g
-
b
ased
p
r
ep
r
o
ce
s
s
in
g
m
ec
h
an
is
m
s
ar
e
ev
alu
ated
to
h
an
d
le
a
h
ig
h
er
n
o
is
e
lev
el
in
th
e
in
p
u
t
f
u
n
d
u
s
im
ag
e.
Ap
ar
t
f
r
o
m
n
o
is
e
elim
in
atio
n
an
d
g
e
o
m
etr
ic
ad
ju
s
tm
en
t,
th
e
s
tu
d
y
also
ev
alu
ates
v
ar
io
u
s
co
n
tr
ast
en
h
an
ce
m
e
n
t
tech
n
iq
u
es su
ch
as p
o
wer
law,
C
L
AHE
,
lo
w
-
lig
h
t c
o
lo
r
en
h
a
n
ce
m
en
t,
m
ath
e
m
atica
l m
o
r
p
h
o
lo
g
y
,
a
n
d
p
ar
ticle
s
war
m
o
p
tim
izatio
n
(
PS
O)
-
b
ased
o
p
tim
izatio
n
to
ad
d
r
ess
lo
w
co
n
tr
ast
f
ac
to
r
s
an
d
u
n
ev
en
illu
m
in
atio
n
.
T
h
is
r
esear
ch
s
tu
d
y
aim
s
to
o
f
f
er
a
n
o
v
el
an
d
v
er
s
atile
im
ag
e
p
r
e
p
r
o
ce
s
s
in
g
m
o
d
el
th
at
au
to
m
a
tes
clin
ical
an
aly
s
is
wh
er
e
tim
e
an
d
ac
cu
r
ac
y
ar
e
c
r
u
cial
f
ac
to
r
s
.
T
h
u
s
,
it
allo
ws
th
e
p
h
y
s
ician
to
c
h
o
o
s
e
p
r
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
f
lex
ib
ilit
y
to
p
er
f
o
r
m
a
s
ig
n
if
ican
t
en
h
an
ce
m
e
n
t
o
v
er
im
a
g
e
f
o
r
ea
r
ly
d
etec
tio
n
o
f
DR
.
T
h
e
co
n
s
id
er
ab
le
co
n
tr
ib
u
tio
n
s
o
f
th
e
p
r
o
p
o
s
ed
r
esear
ch
wo
r
k
a
r
e
as f
o
llo
ws:
−
T
h
e
p
r
o
p
o
s
ed
o
n
-
d
em
an
d
p
r
e
-
p
r
o
ce
s
s
in
g
f
r
am
ewo
r
k
p
r
o
v
id
es
a
p
o
ten
tial
s
o
lu
tio
n
to
ad
d
r
ess
s
ev
er
al
is
s
u
es
o
f
ten
en
co
u
n
te
r
ed
in
th
e
im
ag
e
e
n
h
an
ce
m
e
n
t
p
r
o
ce
s
s
.
I
t
ca
n
also
b
e
u
s
ed
in
ter
ch
an
g
ea
b
ly
d
ep
en
d
i
n
g
o
n
th
e
in
p
u
t im
ag
e'
s
v
is
u
al
ch
ar
ac
ter
is
tics
.
−
T
h
e
ev
alu
atio
n
o
f
th
e
p
r
e
-
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
in
teg
r
ated
in
to
th
e
p
r
o
p
o
s
ed
s
y
s
tem
is
c
ar
r
ied
b
ased
o
n
th
e
o
u
tp
u
t r
esp
o
n
s
e
an
d
th
e
im
ag
e
s
tatis
tic
s
.
−
I
t
als
o
p
r
o
v
id
es
co
s
t
-
ef
f
ec
tiv
en
ess
an
d
tim
e
ef
f
icien
cy
in
clin
ical
an
aly
s
is
.
I
n
ad
d
itio
n
,
p
h
y
s
ician
s
ca
n
co
n
d
u
ct
m
o
r
e
p
r
o
f
o
u
n
d
r
esea
r
ch
o
f
f
u
n
d
u
s
im
ag
in
g
b
y
s
elec
tin
g
d
if
f
er
en
t
p
r
e
-
p
r
o
ce
s
s
in
g
m
ec
h
an
is
m
s
in
teg
r
ated
in
to
t
h
e
p
r
o
p
o
s
ed
s
y
s
tem
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
12
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
5
8
5
-
595
588
−
I
n
ad
d
itio
n
,
th
e
p
r
o
p
o
s
ed
f
r
a
m
ewo
r
k
also
p
r
o
v
id
es
th
e
b
en
ef
it
o
f
o
p
tim
izatio
n
i
n
th
e
en
h
an
ce
m
en
t
p
r
o
ce
s
s
u
s
in
g
a
s
war
m
-
b
ased
s
to
ch
as
tic
ap
p
r
o
ac
h
with
t
h
e
f
lex
i
b
ilit
y
o
f
f
in
e
-
tu
n
in
g
its
p
ar
am
et
er
f
o
r
a
b
etter
an
aly
s
is
o
f
th
e
im
ag
e
b
ased
o
n
th
e
p
ar
ticu
lar
r
eq
u
ir
e
m
en
ts
.
5.
T
H
E
P
RO
P
O
SE
D
AL
G
O
RI
T
H
M
I
n
th
is
s
ec
tio
n
,
th
e
alg
o
r
ith
m
s
ar
e
d
is
cu
s
s
ed
u
s
ed
i
n
th
e
im
p
lem
en
tatio
n
o
f
t
h
e
p
r
o
p
o
s
ed
o
n
-
d
em
an
d
p
r
ep
r
o
ce
s
s
in
g
m
ec
h
an
is
m
.
T
h
e
p
r
im
e
co
n
ce
r
n
b
eh
in
d
th
e
i
m
p
lem
en
tatio
n
o
f
th
is
alg
o
r
ith
m
is
th
at
d
if
f
er
en
t
f
u
n
d
u
s
im
ag
es
co
m
e
with
a
wid
e
v
ar
iety
o
f
p
r
o
b
lem
s
.
T
h
ese
p
r
o
b
lem
s
n
ee
d
to
b
e
ad
d
r
ess
ed
in
th
e
p
r
ep
r
o
ce
s
s
in
g
p
h
ase
as
a
p
r
er
eq
u
is
ite
f
o
r
h
u
m
an
an
aly
s
i
s
an
d
th
e
au
to
m
atio
n
p
r
o
ce
s
s
es
in
th
e
clin
ical
r
esear
ch
to
war
d
s
d
is
ea
s
e
id
en
tific
atio
n
an
d
p
atien
t
tr
ea
tm
en
t.
T
h
er
ef
o
r
e,
th
e
p
r
o
p
o
s
ed
s
y
s
tem
f
o
cu
s
es
o
n
th
r
ee
d
if
f
er
e
n
t
ca
teg
o
r
ies
o
f
p
r
ep
r
o
ce
s
s
in
g
is
s
u
es
an
d
p
r
esen
ts
p
o
ten
tial
s
o
lu
tio
n
s
th
at
w
o
r
k
in
an
in
teg
r
ated
way
to
o
f
f
e
r
a
co
m
p
r
e
h
en
s
iv
e
en
h
an
ce
m
e
n
t o
f
th
e
f
u
n
d
u
s
im
ag
e.
5
.
1
.
Alg
o
rit
hm
f
o
r
g
eo
m
et
ric
a
dj
us
t
m
ent
T
h
e
g
eo
m
etr
ic
ad
ju
s
tm
en
t
p
r
o
ce
s
s
r
ef
er
s
to
r
ec
tify
in
g
d
is
to
r
tio
n
s
ass
o
ciate
d
with
im
ag
e
s
h
ap
e,
o
b
ject
r
ef
lectio
n
,
s
ca
le,
o
r
ie
n
tatio
n
,
an
d
n
o
r
m
aliza
tio
n
.
T
h
e
p
r
o
p
o
s
ed
r
esear
c
h
s
tu
d
y
co
n
s
id
er
s
th
r
ee
g
eo
m
etr
ic
-
r
elate
d
is
s
u
es
;
i)
im
ag
e
u
p
s
ca
lin
g
an
d
d
o
wn
s
ca
li
n
g
ii)
r
eg
io
n
o
f
in
ter
est
m
in
in
g
,
an
d
iii)
g
r
a
y
s
ca
le
n
o
r
m
aliza
tio
n
.
T
h
e
f
u
n
d
u
s
im
ag
e
p
h
o
to
g
r
ap
h
e
r
a
d
o
p
ts
d
if
f
e
r
en
t
f
ield
v
iews
(
Fo
V)
to
ac
h
i
ev
e
f
u
n
d
u
s
im
ag
in
g
.
T
h
u
s
,
th
e
ac
q
u
ir
ed
im
ag
es
a
r
e
v
ar
iab
le
i
n
s
ize
ac
co
r
d
i
n
g
to
d
iag
n
o
s
tic
r
eq
u
ir
e
m
en
ts
.
T
h
e
alg
o
r
ith
m
f
o
r
im
p
lem
e
n
tin
g
th
e
s
ca
lin
g
tech
n
iq
u
e
is
d
escr
ib
ed
as f
o
llo
ws:
Alg
o
r
ith
m
f
o
r
f
u
n
d
u
s
im
a
g
e
up
s
ca
lin
g
an
d
d
o
wn
s
ca
lin
g
Input
: FIm (Input Fundus Image)
Output
: SFIm (Scaled Fundus Image)
Start
1.
[r,c]
f
1
(FIm)
2.
Initialize, Tr, Tc
3.
rR
Tr/r
4.
[cR]
Tc/c
5.
[
Pr
,
Pc
]
=
⌈
(
r
×
rR
/
rR
)
,
⌈
(
c
×
cR
/
cR
)
6.
[r,g,b]
f2(FIm)
7.
r(r,c)
→
r
(
Pr
,
Pc
)
8.
g(r,c)
→
g
(
Pr
,
Pc
)
9.
b(r,c)
→
b
(
Pr
,
Pc
)
10.
Matz
[ ]
TrxTcx3
11.
SFIm
f3(Matz, rgb)
End
I
n
th
e
f
ir
s
t
s
tep
o
f
th
e
alg
o
r
ith
m
,
th
e
s
y
s
tem
u
s
es
a
f
u
n
ctio
n
f
1
(
)
o
v
e
r
th
e
in
p
u
t
im
ag
e
to
g
et
th
e
s
ize
o
f
th
e
in
p
u
t
f
u
n
d
u
s
im
ag
e
FI
M
as
th
e
n
u
m
b
er
o
f
r
o
w
r
an
d
co
lu
m
n
c
(
lin
e1
)
.
I
n
th
e
s
ec
o
n
d
s
tep
,
a
v
ar
ia
b
le
tar
g
et
r
o
w
(
T
r
)
an
d
tar
g
et
co
l
u
m
n
(
T
c)
a
r
e
in
itialized
to
d
ef
in
e
th
e
n
ew
s
ize
o
f
th
e
im
a
g
e
(
lin
e2
)
.
T
o
p
er
f
o
r
m
s
ca
lin
g
,
th
e
alg
o
r
it
h
m
co
m
p
u
t
es
a
s
ca
lin
g
r
ati
o
r
R
,
cR
b
y
d
iv
id
in
g
t
h
e
tar
g
et
im
a
g
e
s
ize
b
y
th
e
s
ize
o
f
th
e
o
r
ig
in
al
im
ag
e
in
a
r
o
w
-
wis
e
an
d
co
lu
m
n
-
wis
e
m
an
n
er
(
L
in
e
3
&
L
in
e
4
)
.
T
h
ese
s
ca
lin
g
r
ati
o
s
ar
e
th
en
u
s
ed
to
co
m
p
u
te
Pr
an
d
Pc'
s
s
ca
l
in
g
p
o
s
itio
n
s
f
o
r
th
e
p
ix
el
r
e
p
licatio
n
p
r
o
ce
s
s
(
L
in
e5
)
.
T
h
e
(
1
)
ex
h
i
b
its
th
e
co
m
p
u
tatio
n
o
f
s
ca
lin
g
p
o
s
itio
n
s
as
:
[
,
]
=
⌈
(
(
×
)
/
)
,
⌈
(
(
×
)
/
)
(
1
)
w
h
er
e
Pr
is
th
e
r
o
w
-
wis
e
s
ca
lin
g
p
o
s
itio
n
an
d
Pc
is
th
e
co
l
u
m
n
-
wis
e
s
ca
lin
g
p
o
s
itio
n
,
⌈
is
th
e
f
u
n
ctio
n
t
h
at
d
en
o
tes
r
o
u
n
d
in
g
to
war
d
s
th
e
n
ea
r
est
in
teg
er
.
T
h
e
alg
o
r
ith
m
d
ef
in
ed
a
f
u
n
ctio
n
f
2
(
)
to
ex
tr
ac
t
th
e
co
lo
r
co
m
p
o
n
en
ts
R
GB
f
r
o
m
th
e
in
p
u
t
im
ag
e
to
p
er
f
o
r
m
a
p
ix
el
r
ep
licatio
n
p
r
o
ce
s
s
in
ea
ch
co
l
o
r
co
m
p
o
n
e
n
t.
T
h
e
co
m
p
u
ted
s
ca
led
r
o
w
p
o
s
itio
n
Pr
an
d
co
lu
m
n
s
ca
led
p
o
s
itio
n
Pc
(
L
in
e6
-
9
)
.
Fu
r
th
e
r
,
a
n
u
ll
m
atr
ix
is
co
n
s
tr
u
cted
w
h
o
s
e
all
e
n
tr
ies
ar
e
ze
r
o
s
with
th
e
s
ize
o
f
th
e
tar
g
et
im
ag
e
(
L
in
e
1
0
)
.
A
n
ew
r
escaled
im
ag
e
is
th
en
o
b
tain
ed
in
th
e
n
e
x
t
s
tep
o
f
t
h
e
alg
o
r
ith
m
u
s
in
g
f
u
n
ctio
n
f
3
(
)
,
wh
ich
p
er
f
o
r
m
s
s
am
p
li
n
g
o
f
all
t
h
r
ee
c
o
lo
r
co
m
p
o
n
en
ts
in
to
th
e
n
u
ll
m
atr
ix
th
at
f
in
ally
co
n
s
tr
u
cts
th
e
n
ew
im
ag
e
(
SF
I
m
)
with
th
e
s
ize
o
f
T
r
an
d
T
c
(
L
in
e1
1
)
.
T
h
e
n
ex
t
o
p
er
atio
n
in
th
e
g
e
o
m
etr
ic
ad
j
u
s
tm
en
t
i
s
th
e
ex
tr
ac
tio
n
o
f
t
h
e
o
b
ject
in
th
e
im
ag
e
u
n
d
er
co
n
s
id
er
atio
n
.
T
h
e
s
tu
d
y
co
n
s
id
er
s
th
is
p
ar
t
a
s
ig
n
if
ican
t
f
u
n
ctio
n
to
in
clu
d
e
in
th
e
p
r
o
p
o
s
ed
s
y
s
tem
.
I
t
f
ac
ilit
ates
th
e
p
h
y
s
ician
to
wo
r
k
o
n
o
n
ly
a
s
p
ec
if
ic
p
ar
t
r
a
th
er
th
an
an
al
y
zi
n
g
th
e
wh
o
le
im
ag
e.
A
cr
o
p
p
in
g
m
ec
h
an
is
m
is
ad
o
p
ted
to
p
er
f
o
r
m
R
o
I
ex
tr
ac
tio
n
,
wh
ich
also
ac
ts
as
a
p
r
im
ar
y
s
eg
m
en
tatio
n
m
ec
h
an
is
m
to
p
er
f
o
r
m
im
ag
e
an
aly
s
is
f
o
r
e
y
e
-
r
elate
d
d
is
ea
s
e
id
en
tific
atio
n
.
T
h
e
n
o
r
m
aliza
tio
n
o
f
in
p
u
t
f
u
n
d
u
s
im
ag
e
is
ca
r
r
ied
o
u
t in
g
r
a
y
s
ca
le
r
ep
r
es
en
tatio
n
u
n
d
er
a
r
an
g
e
o
f
ev
e
n
d
is
tr
ib
u
tio
n
o
f
p
ix
els as
in
(
2
)
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mo
d
ellin
g
o
n
d
ema
n
d
p
r
ep
r
o
c
ess
in
g
fr
a
mewo
r
k
to
w
a
r
d
s
…
(
P
r
a
kru
th
i Ma
n
d
ya
K
r
is
h
n
eg
o
w
d
a
)
589
N
=
(
(
−
p
)
×
255
)
/
(
−
)
(
2
)
wh
er
e
NFI
m
is
th
e
n
o
r
m
alize
d
in
p
u
t,
FIm
is
th
e
n
o
r
m
alize
d
im
ag
ed
,
p
,
an
d
p
is
th
e
m
in
im
u
m
an
d
m
ax
im
u
m
s
ca
le
o
f
t
h
e
in
p
u
t
im
ag
e
FIm
.
T
h
is
also
p
e
r
f
o
r
m
s
a
s
im
ilar
o
p
e
r
atio
n
o
f
d
e
n
o
is
in
g
in
th
e
in
p
u
t
im
ag
e.
5
.
2
.
Alg
o
rit
hm
f
o
r
deno
is
ing
T
h
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
co
n
s
id
er
s
th
r
ee
d
if
f
er
e
n
t
f
ilter
in
g
o
p
er
atio
n
s
,
s
u
ch
as
a
m
ea
n
,
m
ed
ian
,
an
d
co
m
b
in
ed
m
ea
n
-
m
ed
ian
f
ilter
to
d
e
-
n
o
is
e
th
e
im
ag
e
a
n
d
g
en
er
ates
an
en
h
a
n
ce
d
im
a
g
e
as
an
o
u
tp
u
t.
T
h
e
alg
o
r
ith
m
f
o
r
n
o
is
e
f
ilter
in
g
in
th
e
im
ag
e
is
d
escr
ib
ed
as
f
o
llo
ws:
Alg
o
r
ith
m
f
o
r
f
u
n
d
u
s
im
a
g
e
d
en
o
is
in
g
Input
: FI (Input Fundus Image)
Output
: DFIm (Denoised Fundus Image)
Start
1.
FI1
f4(FIm,
W
)
a.
f4
(
)
←
1
/
(
m
x
n
)
∑
FI
m
{
p
,
q
}
N
∈
{
p
,
q
}
2.
FI2
f5(FIm, M) // Secondary level filtering
a.
f5
(
)
←
m
e
dia
n
(
FI
m
(
p
−
i
,
q
−
j
)
,
i
,
j
∈
M
)
3.
FI3
f5(f4(FIw, W)) // Multi
-
level filtering
4.
DFIm
f6(FI3)
End
T
h
e
f
ir
s
t
s
tep
o
f
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
is
m
ain
ly
s
u
b
jecte
d
to
t
h
e
m
ea
n
f
ilter
in
g
p
r
o
ce
s
s
FI1
.
An
av
er
ag
in
g
f
u
n
ctio
n
f
4
(
)
is
u
s
e
d
to
p
e
r
f
o
r
m
th
e
f
ilter
in
g
p
r
o
c
ess
u
s
in
g
lin
ea
r
o
p
er
atio
n
with
F
(
f
ilter
m
ask
=
m
x
n
,
wh
er
e
n
=3
,
an
d
m
=
3
)
.
T
h
e
f
u
n
ctio
n
f
4
(
)
d
eter
m
in
es th
e
m
ea
n
v
alu
e
o
f
t
h
e
FIm
(
p
,
q
)
with
th
e
W
ce
n
ter
ed
at
p
o
in
t p
,
q
p
i
x
el
co
o
r
d
in
ate
o
f
FIm
(
L
in
e1
)
.
T
h
e
f
ilter
ed
im
a
g
e
(
m
FI)
at
p
o
in
t (
p
,
q
)
ca
n
b
e
g
iv
en
as in
(
3
)
:
=
1
/
(
)
∑
{
,
}
∈
{
,
}
(
3
)
w
h
er
e
m
FI
is
th
e
f
ilter
ed
im
ag
e
an
d
W
is
th
e
f
ilter
m
ask
o
r
win
d
o
w
as
a
s
q
u
ar
e
m
atr
ix
o
f
s
ize
m
x
n
,
N
r
ef
er
s
to
th
e
n
eig
h
b
o
r
h
o
o
d
v
alu
es
b
e
lo
n
g
in
g
to
th
e
s
et
o
f
p
ix
el
co
o
r
d
in
ate
p
,
q
in
th
e
s
q
u
ar
e
m
atr
i
x
F.
T
h
is
o
p
er
atio
n
is
ex
ec
u
ted
u
s
in
g
co
n
v
o
lu
tio
n
m
atr
ix
,
wh
e
r
e
ea
ch
co
ef
f
icien
t
co
n
s
id
er
e
d
with
th
e
v
alu
e
o
f
1
/F.
T
h
is
o
p
e
r
atio
n
is
co
m
p
u
tatio
n
ally
e
f
f
icien
t
an
d
ef
f
ec
tiv
e
in
s
m
o
o
th
in
g
th
e
im
ag
e'
s
v
is
u
al
q
u
ality
s
u
f
f
er
in
g
f
r
o
m
t
h
e
b
lu
r
r
in
ess
.
Ho
wev
er
,
if
th
e
in
p
u
t
f
u
n
d
u
s
im
ag
e
is
s
u
b
jecte
d
to
a
h
i
g
h
er
n
o
is
e
o
r
r
an
d
o
m
n
o
is
e
lev
el,
th
e
n
th
e
in
p
u
t
im
a
g
e
p
r
o
ce
s
s
ed
with
m
ea
n
f
ilter
i
n
g
lack
s
t
h
e
ed
g
e
d
etails
in
th
e
r
esu
ltin
g
o
u
tp
u
t.
T
h
e
m
ed
ian
f
ilter
in
g
FI2
is
co
n
s
id
er
ed
in
th
e
p
r
o
p
o
s
ed
s
y
s
tem
to
elim
in
ate
r
an
d
o
m
n
o
is
e
is
s
u
es
in
th
e
in
p
u
t
im
ag
e.
Similar
ly
,
a
m
ed
ia
n
f
ilter
is
ap
p
lied
u
s
in
g
f
u
n
ctio
n
f
5
(
)
,
wh
ich
e
x
ec
u
tes
n
o
n
-
lin
ea
r
o
p
e
r
atio
n
co
n
s
id
er
in
g
c
o
m
p
u
tatio
n
o
f
t
h
e
m
ed
ian
v
alu
e
o
f
th
e
FIm
with
f
ilter
m
ask
M
(
L
i
n
e2
)
.
I
n
th
is
p
r
o
ce
s
s
,
th
e
co
o
r
d
in
ates
p
o
i
n
ts
o
f
a
p
ix
el
in
th
e
m
ask
M
ar
e
ar
r
an
g
e
d
b
a
s
ed
o
n
th
e
g
r
ay
s
ca
le
v
alu
es,
a
n
d
th
e
m
ed
ian
v
alu
e
o
f
M
is
u
s
ed
to
e
lim
in
ate
th
e
n
o
is
y
v
alu
es.
I
t
r
esto
r
es
th
e
n
o
is
e
-
f
r
ee
p
ix
el
v
alu
es
in
th
e
M.
T
h
e
o
u
tp
u
t
im
ag
e
ac
h
iev
ed
b
y
n
o
n
-
lin
ea
r
en
h
a
n
c
em
en
t o
p
er
atio
n
d
escr
ib
e
d
as
(
4
)
:
mdF
I
(
p
,
q
)
=
me
dia
n
(
F
Im
(
p
−
i
,
q
−
j
)
,
i
,
j
∈
M
)
(
4
)
w
h
er
e
m
d
FI
is
th
e
o
u
tp
u
t
d
en
o
is
ed
f
u
n
d
u
s
im
a
g
e,
M
is
th
e
m
ask
o
f
s
ize
n
x
m
,
wh
er
e
m
,
n
=3
x
3
.
T
h
is
tech
n
iq
u
e
is
ef
f
ec
tiv
e
in
th
e
elim
in
atio
n
o
f
r
an
d
o
m
n
o
is
e,
b
lu
r
r
in
ess
in
th
e
f
u
n
d
u
s
im
ag
es.
Ho
wev
er
,
th
e
p
r
o
p
o
s
ed
s
tu
d
y
also
im
p
lem
en
ted
a
c
o
m
b
i
n
ed
a
p
p
r
o
ac
h
o
f
m
ea
n
,
m
ed
i
an
,
an
d
f
ir
s
t
-
o
r
d
e
r
r
ec
u
r
s
iv
e
f
ilter
in
g
ap
p
r
o
ac
h
t
o
p
er
f
o
r
m
a
m
u
lti
-
lev
el
f
ilter
in
g
p
r
o
ce
s
s
to
elim
in
ate
h
ig
h
e
r
-
lev
el
n
o
is
e
in
th
e
in
p
u
t
im
ag
e
(
L
in
e
3
)
.
I
n
th
is
p
r
o
ce
s
s
,
th
e
o
u
tco
m
e
o
b
tain
ed
(
FI1
)
f
r
o
m
th
e
m
ea
n
f
ilter
in
g
a
p
p
r
o
ac
h
is
tak
en
as
in
p
u
t
f
o
r
th
e
m
ed
ian
f
ilter
in
g
ap
p
r
o
ac
h
o
f
m
ask
M
s
ize
5
x
5
.
Ou
tp
u
t
im
ag
e
FI3
is
th
en
p
r
o
ce
s
s
ed
with
a
f
ir
s
t
-
o
r
d
er
r
ec
u
r
s
iv
e
f
ilter
[
2
7
]
.
A
r
e
cu
r
s
iv
e
f
ilter
in
g
f
u
n
ctio
n
f
6
(
)
is
u
s
ed
with
an
in
p
u
t
a
r
g
u
m
en
t
FI3
(
o
u
tp
u
t
o
f
co
m
b
in
ed
m
ea
n
a
n
d
m
e
d
ian
f
i
lter
)
.
5
.
3
.
Alg
o
rit
hm
f
o
r
co
ntr
a
s
t
enha
ncem
ent
V
a
r
io
u
s
im
ag
e
en
h
an
ce
m
e
n
t
tech
n
iq
u
es
im
p
lem
en
te
d
in
th
e
p
r
o
p
o
s
ed
s
y
s
tem
to
co
n
tr
ast
a
d
ju
s
tm
en
t
an
d
b
r
ig
h
tn
ess
p
r
eser
v
atio
n
in
th
e
in
p
u
t
im
ag
e
ar
e
d
is
cu
s
s
ed
in
th
is
s
ec
tio
n
.
Po
wer
-
law
-
b
ased
en
h
an
ce
m
en
t
is
wid
ely
u
s
ed
in
t
h
e
e
x
is
tin
g
liter
atu
r
e
to
im
p
r
o
v
e
th
e
co
n
tr
ast o
f
im
ag
es b
ased
o
n
th
e
g
am
m
a
co
r
r
ec
tio
n
f
ac
to
r
.
Alg
o
r
ith
m
f
o
r
f
u
n
d
u
s
im
a
g
e
C
o
n
tr
ast en
h
an
ce
m
en
t u
s
in
g
p
o
wer
-
law
tr
an
s
f
o
r
m
s
Input
: FI (Input Fundus Image)
Output
: EFI (Enhanced Fundus Image), b2
(brightness EFI), s2 (sharpness EFI)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
12
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
5
8
5
-
595
590
Start
1.
Initialize
2.
EFI
255x
(
FI
(
,
)
/
255
)
1
/
3.
[b1,b2]
f2(EFI(f1(FI))
4.
[s1, s2]
f4(EFI(f3(FI)
5.
Evaluate: output image
6.
/
If
th
e
re
qu
ir
em
en
t
is
to
in
cr
ea
se
th
e
co
nt
ra
st
of
da
rk
an
d
po
or
co
nt
ra
st
fu
nd
us
image:
7.
Check:[b1,b2] and [s1,s2]
8.
Otherwise:
9.
Compute
with
a
differen
t
value
of
0.1:0.98
until
the
suitable
value
of
is
evaluated
End
T
h
e
f
ir
s
t
s
tep
o
f
th
e
alg
o
r
ith
m
is
to
in
itialize
th
e
v
a
r
iab
le
attr
ib
u
te
f
o
r
co
n
tr
ast
co
r
r
ec
tio
n
(
L
in
e1
)
.
T
h
en
,
t
h
e
co
n
tr
ast
co
r
r
ec
tio
n
f
o
r
g
iv
en
im
ag
e
e
n
h
an
ce
m
en
t
is
ca
r
r
ied
u
s
in
g
n
u
m
er
ical
ex
p
r
ess
io
n
m
en
tio
n
ed
in
th
e
alg
o
r
ith
m
,
wh
ich
r
ef
er
s
to
a
n
o
n
-
lin
ea
r
o
p
er
atio
n
th
at
m
ap
s
th
e
u
n
-
ev
e
n
illu
m
in
atio
n
v
alu
e
in
th
e
i
n
p
u
t
im
ag
e
to
th
e
u
n
if
o
r
m
lu
m
in
an
ce
in
th
e
o
u
tp
u
t
s
p
ac
e
u
s
in
g
co
r
r
ec
tio
n
f
ac
t
o
r
f
o
llo
wed
b
y
p
o
wer
-
law
tr
an
s
f
o
r
m
atio
n
(
lin
e
2
)
.
I
n
th
e
n
ex
t
s
tep
(
L
i
n
e3
)
,
th
e
alg
o
r
it
h
m
co
m
p
u
tes
th
e
b
r
ig
h
tn
ess
v
alu
e
b
2
a
n
d
b
1
f
o
r
b
o
th
e
n
h
an
ce
d
im
ag
e
u
s
in
g
f
u
n
ctio
n
f
2
(
)
an
d
in
p
u
t
im
ag
e
u
s
in
g
f
u
n
ctio
n
f
1
(
)
.
Similar
ly
,
th
e
q
u
an
tifie
d
s
h
ar
p
n
ess
s
1
an
d
s
2
ar
e
also
co
m
p
u
ted
f
o
r
b
o
th
in
p
u
t
im
ag
e
an
d
o
u
t
p
u
t
im
ag
e
u
s
in
g
f
u
n
ctio
n
f
3
(
)
an
d
f
4
(
)
r
esp
ec
tiv
ely
f
o
r
th
e
p
ar
ticu
lar
v
alu
e
o
f
th
e
c
o
r
r
ec
tio
n
f
ac
t
o
r
attr
ib
u
te
L
in
e
4
&
5
.
Ho
wev
e
r
,
th
e
r
eq
u
i
r
em
en
t
is
to
in
cr
ea
s
e
th
e
im
ag
e'
s
co
n
t
r
ast;
th
er
ef
o
r
e,
t
h
e
s
y
s
tem
ev
a
lu
ates
b
o
th
q
u
an
tifie
d
s
h
a
r
p
n
e
s
s
an
d
b
r
ig
h
t
n
ess
to
ju
s
tify
th
e
n
ee
d
f
o
r
FI
e
n
h
an
ce
m
en
t u
s
in
g
th
is
tech
n
iq
u
e
(
L
in
e
7
&
8
)
.
H
o
wev
er
,
if
t
h
e
u
s
er
is
n
o
t satis
f
ied
,
th
e
n
th
e
s
am
e
s
tep
s
will
b
e
f
o
llo
we
d
with
d
if
f
e
r
en
t
v
alu
es
o
f
with
in
th
e
r
a
n
g
e
o
f
1
:
0
.
9
8
.
I
n
ad
d
itio
n
,
th
e
v
al
u
e
o
f
ca
n
n
o
t
b
e
co
n
s
id
er
ed
o
n
e
b
e
ca
u
s
e
th
e
p
o
wer
law
tr
an
s
f
o
r
m
atio
n
f
u
n
ctio
n
p
er
f
o
r
m
s
lin
er
o
p
er
atio
n
,
wh
ic
h
m
ea
n
s
n
o
m
ap
p
in
g
o
f
o
u
tp
u
t
s
p
ac
e
with
co
n
tr
ast
co
r
r
ec
te
d
v
alu
es.
T
h
e
e
n
h
an
ce
m
e
n
t
r
eq
u
ir
em
en
t
also
is
ju
s
tifie
d
with
th
e
h
is
to
g
r
a
m
a
n
aly
s
is
.
C
L
AHE
h
as
b
ee
n
an
ex
ten
s
iv
ely
ad
o
p
ted
m
ec
h
a
n
is
m
f
o
r
e
n
h
an
cin
g
t
h
e
co
n
tr
ast
o
f
f
u
n
d
u
s
im
ag
in
g
a
n
d
o
th
e
r
m
ed
ical
im
ag
in
g
m
o
d
alities
.
I
n
C
L
AHE
,
th
e
im
ag
e
is
d
iv
id
ed
in
to
ev
en
ly
d
is
tr
ib
u
ted
b
lo
ck
s
,
an
d
th
e
h
is
to
g
r
am
o
f
ea
c
h
b
l
o
ck
is
d
eter
m
in
ed
.
B
ef
o
r
e
co
m
p
u
tin
g
th
e
g
lo
b
al
h
is
to
g
r
am
an
d
co
n
tr
ast
in
ten
s
ity
,
th
e
lo
ca
l
h
is
to
g
r
am
o
f
ea
ch
b
lo
ck
is
clip
p
ed
u
s
in
g
a
c
lip
p
in
g
lim
it
(
cL
)
d
ef
in
ed
b
ased
o
n
th
e
u
s
er
r
e
q
u
ir
em
en
t.
Fo
r
e
x
a
m
p
le,
f
o
r
in
p
u
t
im
ag
e
FI
with
m
r
o
w
a
n
d
n
co
lu
m
n
,
th
e
cL
is
g
iv
en
b
y
(
5
)
:
=
{
1
(
)
(
)
<
1
(
5
)
w
h
er
e
cF
is
th
e
co
n
tr
ast f
ac
to
r
,
S is
th
e
h
is
to
g
r
am
s
lo
p
e.
T
h
e
clip
p
ed
h
is
to
g
r
am
s
ar
e
n
o
r
m
alize
d
,
a
n
d
th
e
cu
m
u
lativ
e
d
is
tr
ib
u
tio
n
o
f
ea
ch
b
lo
c
k
is
co
m
p
u
te
d
b
ased
o
n
th
e
d
is
tr
ib
u
tio
n
p
a
r
a
m
eter
.
T
h
e
ad
jace
n
t
b
lo
ck
s
ar
e
u
n
ited
u
s
in
g
b
ilin
ea
r
in
ter
p
o
latio
n
to
e
r
ad
icate
u
n
willin
g
b
o
u
n
d
a
r
ies.
T
h
e
ad
v
an
tag
e
o
f
u
s
in
g
C
L
AHE
is
th
at
it
o
v
er
co
m
es
th
e
is
s
u
es
ass
o
ciate
d
with
o
v
er
b
r
ig
h
tn
ess
.
T
h
e
f
u
n
d
u
s
im
ag
es
ar
e
o
f
ten
ca
p
tu
r
ed
in
lo
w
-
lig
h
ten
in
g
co
n
d
itio
n
s
.
As
a
r
esu
lt,
im
ag
es
ar
e
g
en
er
ated
with
lo
w
c
o
n
tr
ast,
h
ig
h
n
o
is
e,
an
d
h
az
in
ess
.
T
h
e
s
tu
d
y
co
n
s
id
er
e
d
a
lo
w
-
lig
h
t
co
lo
r
im
a
g
e
en
h
an
ce
m
e
n
t
tech
n
iq
u
e
to
i
n
cr
ea
s
e
b
r
ig
h
t
n
ess
an
d
s
h
ar
p
n
ess
in
th
e
im
ag
es
ca
p
tu
r
e
d
in
a
lo
w
-
lig
h
tin
g
en
v
ir
o
n
m
en
t.
T
h
e
alg
o
r
ith
m
f
o
r
lo
w
-
lig
h
tin
g
co
lo
r
en
h
a
n
ce
m
en
t is d
is
cu
s
s
ed
as f
o
llo
ws:
Alg
o
r
it
h
m
s
f
o
r
f
u
n
d
u
s
im
ag
e
co
n
tr
ast
en
h
an
ce
m
en
t u
s
in
g
p
o
wer
-
law
tr
an
s
f
o
r
m
Input
: FI (Input Fundus Image)
Output
: EFI
(Enhanced Fundus Image)
Start
1.
Initialize, Inv
2.
Inv1
f1(FI)
3.
E1
f2(Inv)
4.
Inv2
f1(E1)
5.
E2
f2(Inv)
6.
E3
f3(E2)
7.
EFI
f1(E3)
End
I
n
th
e
f
ir
s
t
lin
e
o
f
th
e
alg
o
r
ith
m
,
a
v
ar
iab
le
I
n
v
is
in
itialized
to
c
o
m
p
u
te
an
in
v
e
r
ted
im
a
g
e
(
L
in
e1
)
.
A
f
u
n
ctio
n
f
1
(
)
is
u
s
ed
to
c
o
m
p
u
te
th
e
c
o
m
p
lem
en
t
o
f
i
n
p
u
t
FI,
an
d
th
e
o
u
t
p
u
t
is
s
to
r
ed
as
a
f
ir
s
t
-
lev
el
in
v
er
ted
im
ag
e
(
L
in
e
2
)
.
T
h
e
co
m
p
lem
en
t
o
f
th
e
im
ag
e
is
c
ar
r
ied
o
u
t
u
s
in
g
a
f
o
r
m
u
la,
i.e
.
,
2
5
5
-
FI,
wh
er
e
2
5
5
is
th
e
m
ax
im
u
m
c
o
lo
r
in
ten
s
ity
v
alu
e
.
I
n
th
e
f
o
llo
win
g
p
r
o
c
ess
,
a
f
u
n
ctio
n
f
2
(
)
is
ap
p
lie
d
o
v
er
I
n
v
1
(
i
n
v
er
ted
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mo
d
ellin
g
o
n
d
ema
n
d
p
r
ep
r
o
c
ess
in
g
fr
a
mewo
r
k
to
w
a
r
d
s
…
(
P
r
a
kru
th
i Ma
n
d
ya
K
r
is
h
n
eg
o
w
d
a
)
591
im
ag
e)
to
en
h
an
ce
th
e
im
a
g
e
u
s
in
g
th
e
h
az
e
r
e
m
o
v
al
alg
o
r
i
th
m
(
L
in
e3
)
.
T
o
im
p
r
o
v
e
en
h
an
ce
m
en
t
o
u
t
p
u
t,
a
s
im
ilar
p
r
o
ce
s
s
is
r
ep
ea
ted
to
p
er
f
o
r
m
a
s
ec
o
n
d
ar
y
lev
el
o
f
en
h
an
ce
m
e
n
t.
T
h
er
ef
o
r
e,
a
f
u
n
ctio
n
f
1
(
)
is
ag
ain
ap
p
lied
to
th
e
E
1
,
an
en
h
an
ce
d
im
ag
e
co
m
p
u
te
d
in
th
e
p
r
ev
io
u
s
s
tep
to
co
m
p
u
te
co
m
p
le
m
en
t.
T
h
e
s
ec
o
n
d
ar
y
lev
el
o
f
en
h
an
ce
d
im
ag
e
E
2
i
s
co
m
p
u
t
ed
u
s
in
g
a
s
im
ilar
f
u
n
ctio
n
f
2
(
)
th
at
ca
lls
th
e
h
az
e
r
em
o
v
al
al
g
o
r
ith
m
(
L
in
e4
&
L
in
e
5
)
.
T
h
e
im
p
r
o
v
ed
im
ag
e
E
2
is
f
u
r
th
er
s
u
b
jecte
d
to
th
e
g
u
id
e
d
f
ilter
[
2
8
]
,
u
s
in
g
f
u
n
ctio
n
f
3
(
)
,
wh
ich
p
r
o
v
id
es
ed
g
e
s
m
o
o
th
e
d
im
ag
e
E
3
as
th
e
f
in
al
en
h
an
ce
d
im
ag
e.
Fin
ally
,
th
e
en
h
a
n
ce
d
in
v
er
ted
im
ag
e
is
r
esto
r
ed
in
to
its
o
r
ig
i
n
al
co
l
o
r
s
tate
as
en
h
an
ce
d
f
u
n
d
u
s
i
m
ag
e
(
E
FI
)
en
h
a
n
ce
d
f
u
n
d
u
s
im
ag
e
u
s
in
g
f
u
n
ctio
n
f
1
(
)
.
T
h
e
r
ea
s
o
n
b
eh
in
d
ad
o
p
tin
g
th
e
h
az
e
r
em
o
v
al
-
b
ase
d
lo
w
-
lig
h
t
co
lo
r
f
u
n
d
u
s
im
ag
e
alg
o
r
ith
m
is
to
ex
p
lo
r
e
its
ef
f
ec
tiv
en
ess
to
war
d
s
b
en
ef
itti
n
g
en
h
an
ce
m
e
n
t
o
n
d
ar
k
co
l
o
r
FI
im
ag
es.
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
also
o
f
f
er
s
a
m
ec
h
an
is
m
f
o
r
id
e
n
tify
in
g
o
n
e
o
f
th
e
m
o
s
t
s
ig
n
if
ican
t
f
ea
tu
r
es
k
n
o
wn
as
ex
u
d
ates
u
s
in
g
th
e
m
ath
em
atica
l m
o
r
p
h
o
lo
g
ical
m
ec
h
an
is
m
.
T
h
e
al
g
o
r
i
th
m
ic
s
tep
s
ar
e
g
iv
en
as f
o
llo
ws:
Alg
o
r
ith
m
f
o
r
f
u
n
d
u
s
im
a
g
e
c
o
n
tr
ast en
h
an
ce
m
en
t u
s
in
g
m
o
r
p
h
o
lo
g
ical
o
p
e
r
ato
r
Input
: FI (Input Fundus Image)
Output
: EFI (Enhanced Fundus Image)
Start
1.
[g]
FI(:, :, 2)
2.
sE
f1('disk', 6)
3.
cL
f2(g, sE)
4.
sE
f1('disk', 8)
5.
oP
f3(G, sE)
6.
B
→
f4(oP
-
cL)
End
T
h
e
alg
o
r
ith
m
tak
es
in
p
u
t
f
u
n
d
u
s
im
ag
e
FI
an
d
,
af
ter
a
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
,
p
r
o
v
i
d
es
ex
tr
ac
tio
n
ex
u
d
ates
in
a
c
o
n
to
u
r
p
lo
t.
T
h
e
f
ir
s
t
s
tep
o
f
th
e
al
g
o
r
ith
m
is
to
e
x
tr
ac
t
th
e
g
r
ee
n
'
g
'
co
lo
r
ch
an
n
el
f
r
o
m
FI
to
in
itiate
th
e
m
o
r
p
h
o
lo
g
ical
o
p
e
r
atio
n
(
L
in
e
1
)
.
A
d
is
k
s
h
ap
e
l
ik
e
s
tr
u
ctu
r
in
g
elem
en
t
(
s
E
)
i
s
co
n
s
tr
u
cted
u
s
in
g
f
u
n
ctio
n
f
1
(
)
with
r
a
d
iu
s
len
g
th
o
f
p
ix
els
(
L
i
n
e2
)
t
o
p
er
f
o
r
m
clo
s
in
g
o
p
e
r
atio
n
'
cL
'
f
o
llo
wed
b
y
m
o
r
p
h
o
lo
g
ical
b
o
tto
m
h
at
o
p
e
r
a
tio
n
f
u
n
ctio
n
f
2
(
)
g
iv
en
as
(
6
)
:
⋅
=
(
⨁
)
⊖
(
6
)
w
h
er
e
g
is
th
e
g
r
ee
n
co
lo
r
c
h
a
n
n
el
ex
tr
ac
ted
f
r
o
m
th
e
FI,
s
E
is
th
e
s
tr
u
ctu
r
in
g
elem
en
t,
⋅
clo
s
in
g
o
p
er
ato
r
,
⨁
d
ilatio
n
o
p
er
at
o
r
an
d
⊖
is
th
e
er
o
s
io
n
o
p
er
at
o
r
.
I
n
th
e
b
o
tto
m
h
at
o
p
er
atio
n
,
th
e
o
u
tco
m
e
f
r
o
m
th
e
a
b
o
v
e
n
u
m
er
ical
ex
p
r
ess
io
n
is
s
u
b
tr
ac
t
ed
f
r
o
m
th
e
in
p
u
t
im
ag
e
(
L
in
e
3
)
.
T
h
is
p
r
o
ce
s
s
p
r
o
v
id
es
a
m
ec
h
an
is
m
o
f
v
ess
el
ex
tr
ac
tio
n
o
b
tain
ed
in
th
e
f
o
r
m
o
f
a
lo
w
g
r
ay
s
ca
le
lev
el
o
b
ject.
I
n
th
e
n
ex
t
s
tep
,
th
e
alg
o
r
ith
m
ag
a
in
cr
ea
tes
a
s
tr
u
ctu
r
in
g
elem
en
t
u
s
in
g
a
s
im
ilar
f
u
n
ctio
n
to
in
itiate
o
p
en
in
g
o
p
er
atio
n
f
o
llo
wed
b
y
to
p
h
at
m
o
r
p
h
o
lo
g
ical
(
L
in
e3
an
d
L
i
n
e4
)
.
T
h
e
to
p
th
at
f
u
n
ctio
n
f
3
(
)
ex
tr
ac
ts
a
b
r
i
g
h
t
r
eg
io
n
with
a
h
ig
h
e
r
in
ten
s
ity
lo
ca
lizatio
n
.
T
h
is
o
p
e
r
atio
n
is
g
iv
en
as
(
7
)
:
∘
=
(
⊖
)
⨁
(
7
)
w
h
er
e
∘
is
th
e
m
o
r
p
h
o
l
o
g
ical
o
p
en
in
g
o
p
er
ato
r
.
Fu
r
th
er
,
th
e
alg
o
r
ith
m
p
e
r
f
o
r
m
s
im
ag
e
b
in
ar
izatio
n
o
p
er
ati
o
n
u
s
in
g
f
u
n
ctio
n
f
4
(
)
a
n
d
in
p
u
t
ar
g
u
m
en
t
o
f
th
e
d
if
f
er
en
ce
b
etwe
en
to
p
h
at
an
d
b
o
tto
m
h
at
v
alu
es.
T
h
i
s
m
ea
n
s
th
e
o
u
tc
o
m
e
ac
h
iev
e
d
b
y
th
e
b
o
tto
m
h
at
is
s
u
b
tr
ac
ted
f
r
o
m
th
e
o
u
tco
m
e
ac
h
iev
ed
b
y
th
e
to
p
h
at
p
r
o
ce
s
s
to
ex
tr
ac
t
p
r
ec
is
e
ex
u
d
ates
u
s
in
g
a
co
n
to
u
r
p
lo
t
(
L
in
e6
-
7
)
.
T
h
e
s
tu
d
y
ex
p
lo
r
es
th
e
ef
f
ec
tiv
en
ess
o
f
ap
p
ly
in
g
s
war
m
-
b
ased
o
p
tim
izatio
n
in
th
e
en
h
an
ce
m
e
n
t
o
f
th
e
f
u
n
d
u
s
i
m
ag
e.
T
h
e
o
p
ti
m
izatio
n
m
et
h
o
d
is
b
ased
o
n
th
e
p
o
p
u
lati
o
n
-
o
r
ie
n
ted
s
ea
r
ch
alg
o
r
ith
m
.
T
h
e
s
o
lu
tio
n
s
p
ac
e
is
ca
lled
a
p
ar
ticle,
an
d
ea
c
h
p
ar
ticle
h
as
its
f
itn
ess
v
alu
e
Fv
.
T
h
e
p
ar
ticle
f
r
o
m
th
e
p
r
o
b
lem
s
ea
r
ch
es
th
e
s
o
lu
tio
n
s
p
ac
e
b
y
o
p
tim
a
in
th
e
iter
atio
n
.
E
ac
h
s
o
lu
tio
n
i
s
u
p
d
a
ted
with
th
e
lo
ca
l
b
est
(
Pb
)
an
d
g
lo
b
al
b
est
(
g
b
)
in
ea
ch
iter
atio
n
,
a
n
d
f
u
r
t
h
er
,
it
u
p
d
ated
th
eir
v
elo
city
an
d
p
o
s
itio
n
.
T
h
is
p
r
o
ce
s
s
co
n
tin
u
es
u
n
til
th
e
r
eq
u
ir
em
en
t
is
f
u
lly
s
atis
f
ied
o
r
an
y
er
r
o
r
o
cc
u
r
s
.
T
h
e
alg
o
r
ith
m
s
tep
s
ar
e
m
en
tio
n
ed
f
o
r
p
r
o
p
o
s
ed
en
h
an
ce
m
en
t o
p
tim
izatio
n
.
T
h
e
alg
o
r
ith
m
tak
es
th
e
in
p
u
t
v
alu
e
as
FI
(
f
u
n
d
u
s
im
ag
e)
an
d
g
en
er
ates
E
FI
af
ter
ex
e
cu
tin
g
an
o
p
tim
izatio
n
alg
o
r
ith
m
(
en
h
a
n
ce
d
f
u
n
d
u
s
im
ag
e)
.
First,
th
e
alg
o
r
ith
m
in
itializes
th
e
v
a
r
iab
les
as
m
ax
-
I
ter
,
I
m
ax
,
I
m
in
(
m
a
x
im
u
m
a
n
d
m
in
im
u
m
I
n
er
tial
weig
h
t)
,
v
1
,
a
n
d
v
2
as
ac
ce
ler
atio
n
co
ef
f
icien
ts
,
Ps
ize
p
o
p
u
latio
n
s
ize
(
L
in
e1
)
.
T
h
en
,
th
e
em
p
t
y
v
ec
to
r
co
n
s
tr
u
ctio
n
is
ca
r
r
ied
o
u
t
to
s
to
r
e
Fit
n
ess
v
alu
e
Fv
,
lo
ca
l
b
est
Pb
,
an
d
g
lo
b
al
b
est
g
b
p
o
s
itio
n
o
f
th
e
p
a
r
ticles
(
L
in
e2
)
.
T
h
e
o
v
er
all
co
m
p
u
tatio
n
is
ca
r
r
ied
o
n
th
e
g
r
ay
s
ca
le
im
ag
e
g
FI.
T
h
e
s
ize
o
f
th
e
o
r
ig
in
al
im
ag
e
r
o
w
'
r
,
'
co
lu
m
n
'
c'
is
co
m
p
u
ted
as
a
r
e
f
er
en
ce
v
ar
iab
le
f
o
r
th
e
o
th
er
co
m
p
u
tatio
n
p
r
o
ce
s
s
(
L
in
e3
)
.
Fo
r
ea
c
h
iter
,
co
m
p
u
tatio
n
o
f
v
ar
iab
le
I
is
ca
r
r
ied
o
u
t
to
d
eter
m
in
e
th
e
o
p
tim
al
s
et
f
o
r
p
ar
am
ete
r
s
in
itialized
.
I
t
also
m
ea
n
s
th
at
t
h
e
p
ar
am
ete
r
s
ar
e
s
u
b
jecte
d
t
o
ea
ch
p
ar
ticle
o
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
12
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
5
8
5
-
595
592
s
o
lu
tio
n
s
p
ac
e
(
L
in
e4
&
L
in
e
5
)
.
E
ac
h
p
a
r
ticle
s
ize
co
m
p
u
tes
en
h
an
ce
d
FI
u
s
in
g
f
u
n
ctio
n
f
x
2
(
)
with
an
in
p
u
t
ar
g
u
m
en
t
o
f
g
FI,
an
d
in
itial
p
ar
ticle
p
o
s
itio
n
.
Her
e
th
e
f
u
n
c
tio
n
f
x
2
(
)
r
ef
er
s
to
a
tr
an
s
f
o
r
m
atio
n
o
p
er
atio
n
to
en
h
an
ce
th
e
in
p
u
t
im
ag
e.
I
n
th
e
n
e
x
t
s
tep
o
f
th
e
alg
o
r
ith
m
,
d
is
cr
ete
f
u
n
ctio
n
f
x
3
(
)
with
a
n
in
p
u
t
ar
g
u
m
en
t
o
f
th
e
en
h
a
n
ce
d
im
ag
e,
m
,
an
d
n
is
co
n
s
id
er
ed
f
o
r
co
m
p
u
tin
g
f
itn
e
s
s
v
alu
e
Fv
.
T
h
e
m
ax
im
u
m
v
al
u
e
in
th
e
Fv
m
ax
g
ets
co
m
p
u
ted
.
T
h
e
v
ar
i
ab
le
g
b
is
co
m
p
u
ted
as
th
e
m
ax
im
u
m
v
alu
e
o
f
th
e
Pb
(
L
in
e6
)
.
T
h
e
u
p
d
atin
g
p
r
o
ce
s
s
f
o
r
th
e
o
f
th
e
p
ar
ticle
n
ew
p
o
s
itio
n
P(x
)
an
d
v
elo
ci
ty
P(v
)
is
c
o
n
tin
u
e
d
till
th
e
r
e
q
u
ir
ed
cr
iter
ia
ar
e
m
at
ch
ed
(
L
in
e
7
to
9
)
,
wh
e
r
e
w
is
th
e
I
n
er
tial w
eig
h
t I
a
n
d
Pc
p
r
ev
io
u
s
v
elo
city
an
d
Pl is
th
e
last
lo
ca
tio
n
.
Alg
o
r
ith
m
: Swa
r
m
b
ased
s
to
c
h
asti
c
o
p
tim
izatio
n
ap
p
r
o
ac
h
f
o
r
im
ag
e
e
n
h
an
ce
m
e
n
t
Input: FI (Fundus Image)
Output: FIE (Enhanced Fundus image)
Start
1.
Init,
Max
-
Iteration, Imax, Imin, v1, v2, Psize
2.
Construct: empty vector
[
⃗
⃗
⃗
⃗
⃗
,
⃗
⃗
⃗
⃗
⃗
,
⃗
⃗
⃗
⃗
]
3.
Convert rgb IF
→
gIF(mxn)
f
x1(IF, m x n x D)
4.
Particle Initialization
5.
For i=
1 to Max
-
iteration
a.
Compute:
I
−
(
−
)
×
(
−
)
b.
For j=
1 to Psize
i.
Compute FIE:
x2
(
gFI
,
si
z
e
)
ii.
Compute: Fv
→
3
(
FI
E
,
n
,
m
)
iii.
Fvmax
4
(
Fv
)
iv.
Pb=P(x)
→
c
he
c
k
:
fv
<
(
)
v.
Otherwise:
=
End
End
6.
Update gb
f
4(Pb)
7.
For each P
a.
Compute: P(v)
b.
Compute:
P(x)
8.
Continue until the required criteria are not met.
9.
Stop
End
6.
RE
SU
L
T
S
A
ND
D
I
SCU
SS
I
O
N
T
h
e
im
p
lem
en
tatio
n
o
f
th
e
p
r
o
p
o
s
ed
o
n
-
d
e
m
an
d
p
r
ep
r
o
ce
s
s
in
g
f
r
am
ewo
r
k
is
ca
r
r
ied
o
u
t
o
n
th
e
n
u
m
er
ical
c
o
m
p
u
tin
g
to
o
l
MA
T
L
AB
.
T
h
is
s
ec
tio
n
d
i
s
cu
s
s
es
th
e
o
u
tco
m
es
ac
h
i
ev
ed
b
y
d
if
f
er
e
n
t
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es
f
o
r
f
u
n
d
u
s
im
ag
e
d
e
n
o
is
in
g
a
n
d
en
h
an
ce
m
e
n
t.
E
ac
h
m
eth
o
d
is
ev
alu
ated
o
v
er
s
i
x
d
if
f
er
en
t
f
u
n
d
u
s
im
a
g
es,
wh
er
e
o
n
e
is
a
d
a
r
k
f
u
n
d
u
s
im
ag
e,
an
d
th
e
s
ec
o
n
d
o
n
e
is
a
clea
r
f
u
n
d
u
s
im
ag
e.
T
h
e
s
elec
tio
n
o
f
e
n
h
an
ce
d
im
a
g
es
will
b
e
b
ased
o
n
th
e
v
is
u
al
o
u
tco
m
e
an
d
q
u
an
titativ
e
a
n
aly
s
is
c
o
n
s
i
d
e
r
i
n
g
P
S
N
R
a
n
d
q
u
a
n
t
i
f
i
e
d
b
r
i
g
h
t
n
e
s
s
.
I
t
c
a
n
b
e
s
e
e
n
t
h
a
t
d
i
f
f
e
r
e
n
t
p
r
e
p
r
o
c
e
s
s
i
n
g
o
p
e
r
a
t
i
o
n
p
r
o
v
i
d
e
s
d
i
f
f
e
r
e
n
t
o
u
t
c
o
m
e
s
i
n
t
e
r
m
s
o
f
e
n
h
a
n
c
e
d
i
m
a
g
e
.
T
h
e
q
u
a
n
t
i
f
i
e
d
a
n
a
l
y
s
i
s
i
n
t
e
r
m
s
o
f
P
S
N
R
a
n
d
B
r
i
g
h
t
n
e
s
s
i
s
s
h
o
w
n
i
n
F
i
g
u
r
e
s
2
(
a)
an
d
(
b
)
an
d
Fig
u
r
e
s
3
(
a)
an
d
(
b
)
.
Fig
u
r
e
2
(
a)
th
e
an
aly
s
is
o
f
en
h
an
ce
m
en
t
tech
n
iq
u
es
is
ca
r
r
ied
o
u
t
in
PS
NR
,
wh
ich
s
h
o
ws
b
etter
p
er
f
o
r
m
an
ce
ac
h
iev
ed
b
y
th
e
C
L
AH
E
tech
n
i
q
u
e
an
d
o
th
er
PS
O
o
p
tim
izatio
n
tech
n
iq
u
es.
Fo
r
lin
ea
r
ev
alu
atio
n
,
th
e
clip
lim
it
(
0
.
0
2
)
a
n
d
g
am
m
a
c
o
r
r
ec
ti
o
n
(
0
.
8
9
)
v
al
u
es
ar
e
c
o
n
s
id
er
ed
f
ix
e
d
f
o
r
all
s
ix
im
ag
es.
Fo
r
t
h
e
in
p
u
t
f
u
n
d
u
s
im
ag
es
(
1
,
2
,
5
,
a
n
d
6
)
,
g
a
m
m
a
co
r
r
ec
tio
n
h
as
s
h
o
wn
h
ig
h
er
PS
NR
.
I
n
th
e
o
v
er
all
ev
alu
atio
n
p
r
o
ce
s
s
,
th
e
lo
w
-
lig
h
t
en
h
a
n
ce
m
en
t
tech
n
iq
u
e
ex
h
ib
ited
lo
w
PNSR
co
m
p
ar
ed
to
all
o
t
h
er
m
eth
o
d
s
.
T
h
e
id
ea
o
f
c
o
n
s
id
e
r
in
g
a
l
o
w
-
lig
h
t
e
n
h
an
ce
m
e
n
t
im
ag
e
is
to
en
h
an
ce
t
h
e
v
is
ib
i
lity
o
f
d
a
r
k
f
u
n
d
u
s
im
ag
es.
T
h
e
p
er
f
o
r
m
an
ce
e
v
alu
atio
n
ex
h
ib
ited
in
te
r
m
s
o
f
b
r
ig
h
tn
ess
s
h
o
ws
th
at
lo
w
-
lig
h
t
en
h
an
ce
m
e
n
t
tech
n
iq
u
es
ex
h
ib
it
co
n
s
is
ten
tly
h
ig
h
er
p
er
f
o
r
m
an
ce
.
T
h
is
ev
id
en
t
th
e
r
eq
u
ir
em
en
t
o
f
lo
w
-
lig
h
t
en
h
an
ce
m
en
t
tech
n
iq
u
es in
th
e
clin
ical
a
n
al
y
s
is
f
o
r
th
e
d
ar
k
f
u
n
d
u
s
im
a
g
e
s
.
Fig
u
r
e
3
(
a)
s
h
o
ws
s
i
m
ilar
p
er
f
o
r
m
a
n
ce
b
y
th
e
m
ea
n
f
ilt
er
in
g
ap
p
r
o
ac
h
a
n
d
co
m
b
in
e
d
f
ilter
in
g
ap
p
r
o
ac
h
.
Ho
wev
er
,
th
e
p
e
r
f
o
r
m
an
ce
ex
h
ib
ited
b
y
th
e
m
ed
i
an
f
ilter
in
b
o
th
an
aly
s
es
PS
N
R
an
d
B
r
ig
h
tn
ess
i
s
a
litt
le
le
s
s
co
m
p
ar
ed
to
o
th
er
tech
n
iq
u
es.
T
h
e
r
ea
s
o
n
ca
n
b
e
m
u
ltip
le.
T
h
e
f
i
r
s
t o
n
e
is
th
e
v
is
u
al
ch
ar
ac
ter
is
tic
o
f
in
p
u
t
im
ag
es,
f
ilter
s
ize,
n
o
is
e
n
atu
r
e,
an
d
n
o
is
e
v
ar
ian
ce
in
th
e
in
p
u
t
im
ag
e.
T
h
e
s
t
u
d
y
h
as
co
n
s
id
er
ed
ad
d
in
g
s
o
m
e
ad
d
itiv
e
Ga
u
s
s
ian
n
o
is
e
in
th
e
in
p
u
t
im
ag
e
f
o
r
th
e
p
r
ac
tical
ev
alu
atio
n
,
wh
ic
h
is
th
en
f
ilter
ed
b
y
in
tr
o
d
u
ce
d
f
ilter
in
g
tech
n
i
q
u
es
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
p
r
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es c
an
b
e
im
p
r
o
v
ed
b
y
f
i
n
e
-
tu
n
in
g
th
eir
p
ar
am
eter
s
lik
e
in
th
e
ca
s
e
o
f
g
am
m
a
co
r
r
ec
tio
n
(
g
am
m
a
v
alu
e
v
ar
iatio
n
)
,
C
L
AHE
(
i.e
.
,
clip
-
lim
it
ad
ju
s
tm
en
t)
,
an
d
PS
O
(
p
ar
ticle
p
ar
am
eter
s
)
.
T
h
e
u
s
ag
e
o
f
im
ag
e
s
ca
lin
g
,
cr
o
p
p
in
g
,
an
d
n
o
r
m
aliza
tio
n
p
r
o
v
id
es
e
x
tr
a
ass
is
tan
ce
f
o
r
ef
f
ec
tiv
e
en
h
an
ce
m
e
n
t
o
p
er
atio
n
s
.
I
m
a
g
e
s
ca
lin
g
ca
n
b
e
d
o
n
e
f
o
r
r
e
d
u
cin
g
o
r
in
cr
ea
s
in
g
th
e
r
eso
l
u
tio
n
o
f
th
e
im
a
g
es
co
n
ce
r
n
in
g
co
m
p
u
tatio
n
al
c
o
m
p
lex
ity
.
T
h
e
im
ag
e
cr
o
p
p
i
n
g
ca
n
also
p
r
o
v
id
e
a
b
etter
ap
p
r
o
ac
h
to
v
is
u
alizin
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mo
d
ellin
g
o
n
d
ema
n
d
p
r
ep
r
o
c
ess
in
g
fr
a
mewo
r
k
to
w
a
r
d
s
…
(
P
r
a
kru
th
i Ma
n
d
ya
K
r
is
h
n
eg
o
w
d
a
)
593
th
e
s
p
ec
if
ic
p
ar
t
o
f
t
h
e
in
p
u
t
im
ag
e.
T
h
e
n
o
r
m
aliza
tio
n
m
ec
h
an
is
m
m
ak
es
th
e
d
is
tr
ib
u
t
io
n
o
f
im
a
g
e
p
ix
el
in
ten
s
ity
in
to
a
f
a
v
o
r
ab
le
r
an
g
e.
A
m
o
r
p
h
o
lo
g
ical
ap
p
r
o
ac
h
i
s
co
n
s
id
er
ed
to
d
etec
t
th
e
ex
u
d
ates'
f
ea
t
u
r
es
f
r
o
m
th
e
in
p
u
t
f
u
n
d
u
s
im
a
g
e
as
a
s
e
co
n
d
ar
y
p
r
ep
r
o
ce
s
s
in
g
s
tep
.
I
t
ca
n
b
e
r
e
g
ar
d
ed
as
th
e
p
r
im
ar
y
s
tep
to
war
d
s
t
h
e
im
ag
e
s
eg
m
en
tatio
n
p
r
o
ce
s
s
.
T
h
e
ex
t
r
ac
tio
n
o
f
ex
u
d
ates
is
an
ef
f
e
ctiv
e
b
io
-
m
ar
k
er
to
war
d
s
th
e
d
etec
tio
n
o
f
DR
.
T
h
e
o
u
tco
m
e
o
f
th
e
m
o
r
p
h
o
l
o
g
ical
-
b
ased
p
r
ep
r
o
ce
s
s
in
g
o
p
er
atio
n
is
d
em
o
n
s
tr
ated
i
n
Fig
u
r
e
4
.
(
a)
(
b
)
Fig
u
r
e
2
.
E
v
alu
atio
n
o
f
en
h
an
ce
m
en
t te
ch
n
iq
u
e
in
ter
m
s
o
f
(
a)
PS
NR
,
(
b
)
b
r
ig
h
t
n
ess
(
a)
(
b
)
Fig
u
r
e
3
.
E
v
alu
atio
n
o
f
f
ilter
in
g
t
ec
h
n
iq
u
e
in
ter
m
s
o
f
b
r
i
g
h
t
n
ess
:
(
a)
n
u
m
b
e
r
v
s
.
PS
NR
(
d
b
)
,
(
b
)
n
u
m
b
er
o
f
in
p
u
t im
a
g
es v
s
.
b
r
ig
h
t
n
ess
(
cd
/m
2
)
(
a)
(
b
)
(
c)
Fig
u
r
e
4
.
Prim
ar
y
s
eg
m
en
tatio
n
o
f
e
x
u
d
ate
s
u
s
in
g
m
ath
em
ati
ca
l m
o
r
p
h
o
lo
g
y
:
(
a)
m
o
r
p
h
o
lo
g
ical
b
o
tto
m
h
at
,
(
b
)
m
o
r
p
h
o
lo
g
ical
to
p
h
at,
(
c
)
ex
u
d
ates d
etec
tio
n
B
ased
o
n
th
e
p
er
f
o
r
m
a
n
ce
ev
alu
atio
n
,
th
e
p
r
o
p
o
s
ed
s
tu
d
y
co
n
clu
d
ed
th
at
s
elec
tin
g
s
u
itab
le
en
h
an
ce
m
e
n
t
tech
n
iq
u
es
d
ep
en
d
s
o
n
th
e
o
u
tp
u
t
im
ag
e
v
is
u
al
r
esp
o
n
s
e,
s
tatis
tics
,
a
n
d
q
u
ality
m
etr
ic.
T
h
er
ef
o
r
e,
it
ca
n
b
e
s
aid
t
h
at
ea
ch
p
r
ep
r
o
ce
s
s
in
g
ap
p
r
o
ac
h
in
tr
o
d
u
ce
d
in
th
e
p
r
o
p
o
s
ed
s
y
s
tem
is
a
co
m
p
lem
en
t
to
ea
ch
o
t
h
er
d
e
p
en
d
in
g
o
n
th
e
d
if
f
er
en
t
p
r
e
p
r
o
ce
s
s
in
g
is
s
u
es
an
d
en
h
an
ce
m
en
t
r
eq
u
ir
e
d
.
T
h
e
s
ig
n
if
ican
t
co
n
tr
ib
u
tio
n
o
f
th
i
s
s
tu
d
y
is
th
at
it
o
f
f
er
s
th
e
e
s
s
en
tials
o
f
f
u
n
d
u
s
im
ag
e
q
u
ality
im
p
r
o
v
em
e
n
t,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
12
,
No
.
1
,
Feb
r
u
ar
y
20
22
:
5
8
5
-
595
594
wh
ich
is
o
n
-
d
em
a
n
d
an
d
f
lex
i
b
le
in
an
ac
tiv
e
e
n
v
ir
o
n
m
en
t t
o
p
er
f
o
r
m
co
m
p
r
e
h
en
s
iv
e
en
h
an
ce
m
en
t o
v
e
r
in
p
u
t
im
ag
e.
T
h
er
ef
o
r
e,
a
c
o
llectio
n
o
f
p
r
e
p
r
o
ce
s
s
in
g
p
ac
k
ag
es
with
an
in
ter
ac
tiv
e
u
s
er
in
t
er
f
ac
e
is
p
r
o
p
o
s
ed
,
wh
ich
o
f
f
e
r
s
a
wid
e
r
an
g
e
o
f
f
le
x
ib
ilit
y
to
en
h
a
n
ce
th
e
f
u
n
d
u
s
im
ag
es.
7.
CO
NCLU
SI
O
N
I
n
th
is
p
ap
er
,
a
n
o
n
-
d
em
an
d
p
r
ep
r
o
ce
s
s
in
g
f
r
am
ew
o
r
k
is
m
o
d
eled
to
ad
d
r
ess
d
if
f
er
en
t
en
h
an
ce
m
en
t
is
s
u
es a
n
d
r
an
d
o
m
n
o
is
e
ass
o
ciate
d
with
th
e
f
u
n
d
u
s
im
ag
e.
T
h
e
p
r
o
p
o
s
ed
p
r
ep
r
o
ce
s
s
in
g
f
r
a
m
ewo
r
k
in
teg
r
ated
va
r
io
u
s
tech
n
iq
u
es
t
o
a
d
d
r
e
s
s
o
th
er
p
r
o
b
lem
s
ca
te
g
o
r
iz
ed
in
to
th
r
ee
d
is
tin
ct
p
a
r
ts
:
g
eo
m
et
r
ic
-
r
elate
d
p
r
o
b
lem
s
,
r
a
n
d
o
m
n
o
is
e
an
d
h
ig
h
-
lev
el
n
o
is
e,
an
d
en
h
a
n
ce
m
en
t
-
r
elate
d
p
r
o
b
lem
s
.
T
h
e
s
tu
d
y
im
p
lem
en
te
d
an
im
ag
e
s
ca
lin
g
tech
n
i
q
u
e,
R
o
I
ex
tr
ac
tio
n
,
a
n
d
im
a
g
e
n
o
r
m
ali
za
tio
n
to
a
d
d
r
ess
g
eo
m
et
r
ic
is
s
u
es
in
th
e
im
ag
e.
Similar
ly
,
d
if
f
er
e
n
t
f
ilter
in
g
tech
n
iq
u
es
ar
e
ad
o
p
ted
t
o
a
d
d
r
ess
r
an
d
o
m
a
n
d
h
ig
h
-
le
v
el
n
o
is
e.
C
o
n
tr
ast
en
h
an
ce
m
e
n
t
is
im
p
r
o
v
ed
u
s
in
g
g
am
m
a
co
r
r
ec
tio
n
,
h
is
to
g
r
am
an
aly
s
is
,
lo
w
lig
h
t
co
lo
r
en
h
a
n
ce
m
en
t,
m
o
r
p
h
o
lo
g
ical
o
p
e
r
ato
r
,
an
d
s
war
m
o
p
tim
izatio
n
-
b
ased
en
h
an
ce
m
en
t.
T
h
e
r
ef
o
r
e,
th
e
p
r
o
p
o
s
ed
s
y
s
tem
ca
n
o
f
f
er
a
f
lex
i
b
le
s
o
lu
tio
n
to
o
v
e
r
co
m
e
th
e
lim
itatio
n
s
ass
o
ciat
ed
with
th
e
s
in
g
le
im
ag
e
en
h
a
n
ce
m
en
t
tech
n
iq
u
e.
T
h
e
m
ain
o
b
jectiv
e
o
f
th
is
wo
r
k
is
to
in
tr
o
d
u
c
e
a
u
n
iv
er
s
al
in
ter
f
ac
e
s
y
s
tem
th
at
s
u
p
p
o
r
ts
o
n
-
d
em
a
n
d
p
r
ep
r
o
ce
s
s
in
g
o
p
er
atio
n
s
with
an
o
p
tio
n
o
f
a
wid
e
r
an
g
e
o
f
p
r
e
p
r
o
ce
s
s
in
g
m
ec
h
an
is
m
s
.
T
h
e
u
s
er
ca
n
ef
f
icien
tly
p
e
r
f
o
r
m
ap
p
r
o
p
r
iat
e
en
h
an
ce
m
en
t
o
v
er
an
im
ag
e
b
ased
o
n
e
v
alu
atin
g
a
p
ar
ticu
lar
m
ed
ical
im
ag
e
with
d
if
f
er
en
t p
r
e
p
r
o
ce
s
s
in
g
o
p
tio
n
s
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
ea
ch
tech
n
iq
u
e
is
v
alid
ated
b
ase
d
o
n
q
u
alitativ
e
an
d
q
u
an
titativ
e
an
aly
s
is
.
T
h
e
p
r
o
p
o
s
ed
s
tu
d
y
ca
n
b
e
ex
te
n
d
ed
t
o
p
er
f
o
r
m
s
eg
m
en
tatio
n
an
d
d
iab
atic
r
etin
o
p
ath
y
d
is
ea
s
e
class
if
icatio
n
in
f
u
tu
r
e
r
esea
r
ch
wo
r
k
.
RE
F
E
R
E
NC
E
S
[
1
]
M
.
W
.
S
t
e
w
a
r
t
,
D
i
a
b
e
t
i
c
re
t
i
n
o
p
a
t
h
y
:
C
u
rre
n
t
p
h
a
rm
a
c
o
l
o
g
i
c
t
r
e
a
t
m
e
n
t
a
n
d
e
m
e
rg
i
n
g
st
r
a
t
e
g
i
e
s
,
S
p
r
i
n
g
e
r
,
A
D
I
S
,
2
0
1
7
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
10
-
3
5
0
9
-
8.
[
2
]
A
.
R
.
S
h
a
h
,
a
n
d
T.
W
.
G
a
r
d
n
e
r
,
“
D
i
a
b
e
t
i
c
R
e
t
i
n
o
p
a
t
h
y
:
R
e
s
e
a
r
c
h
t
o
c
l
i
n
i
c
a
l
p
r
a
c
t
i
c
e
,
”
C
l
i
n
i
c
a
l
d
i
a
b
e
t
e
s
a
n
d
e
n
d
o
c
r
i
n
o
l
o
g
y
,
v
o
l
.
3
,
n
o
.
1
,
p
p
.
1
-
7
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
8
6
/
s
4
0
8
4
2
-
0
1
7
-
0
0
4
7
-
y.
[
3
]
J.
A
mi
n
,
M
.
S
h
a
r
i
f
,
a
n
d
M
.
Y
a
smi
n
,
“
A
r
e
v
i
e
w
o
n
r
e
c
e
n
t
d
e
v
e
l
o
p
m
e
n
t
s
f
o
r
d
e
t
e
c
t
i
o
n
o
f
d
i
a
b
e
t
i
c
R
e
t
i
n
o
p
a
t
h
y
,
”
S
c
i
e
n
t
i
f
i
c
a
,
v
o
l
.
2
0
1
6
,
2
0
1
6
,
A
r
t
.
n
o
.
6
8
3
8
9
7
6
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
1
6
/
6
8
3
8
9
7
6
.
[
4
]
B
.
J.
F
e
n
n
e
r
,
R
.
L
.
W
o
n
g
,
W
.
C
.
L
a
m,
G
.
S
.
Ta
n
,
a
n
d
G
.
C
.
C
h
e
u
n
g
,
“
A
d
v
a
n
c
e
s
i
n
r
e
t
i
n
a
l
i
m
a
g
i
n
g
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s
i
n
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
scr
e
e
n
i
n
g
:
a
r
e
v
i
e
w
,
”
O
p
h
t
h
a
l
m
o
l
o
g
y
a
n
d
t
h
e
r
a
p
y
,
v
o
l
.
1
7
,
n
o
.
2
,
p
p
.
3
3
3
-
3
4
6
,
2
0
1
8
,
d
o
i
:
1
0
.
1
0
0
7
/
s
4
0
1
2
3
-
0
1
8
-
0
1
5
3
-
7.
[
5
]
A
.
R
.
S
o
s
a
l
e
,
“
S
c
r
e
e
n
i
n
g
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
i
s
t
h
e
u
se
o
f
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
a
n
d
c
o
st
-
e
f
f
e
c
t
i
v
e
f
u
n
d
u
s
i
ma
g
i
n
g
t
h
e
a
n
sw
e
r
?
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
D
i
a
b
e
t
e
s i
n
D
e
v
e
l
o
p
i
n
g
C
o
u
n
t
ri
e
s
,
v
o
l
.
3
9
,
p
p
.
1
-
3
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
3
4
1
0
-
0
1
9
-
0
0
7
2
9
-
y.
[
6
]
A
.
R
a
j
,
A
.
K
.
Ti
w
a
r
i
,
a
n
d
M
.
G
.
M
a
r
t
i
n
i
,
“
F
u
n
d
u
s
i
m
a
g
e
q
u
a
l
i
t
y
a
ssess
m
e
n
t
:
s
u
r
v
e
y
,
c
h
a
l
l
e
n
g
e
s,
a
n
d
f
u
t
u
r
e
sc
o
p
e
,
”
I
ET
I
m
a
g
e
Pro
c
e
ssi
n
g
,
v
o
l
.
1
3
,
n
o
.
8
,
p
p
.
1
2
1
1
-
1
2
2
4
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
4
9
/
i
e
t
-
i
p
r
.
2
0
1
8
.
6
2
1
2
.
[
7
]
F
.
S
h
a
o
,
Y
.
Y
a
n
g
,
Q
.
Ji
a
n
g
,
G
.
J
i
a
n
g
,
a
n
d
Y
.
H
o
,
“
A
u
t
o
m
a
t
e
d
q
u
a
l
i
t
y
a
ss
e
ssm
e
n
t
o
f
f
u
n
d
u
s
i
ma
g
e
s
v
i
a
a
n
a
l
y
s
i
s
o
f
i
l
l
u
m
i
n
a
t
i
o
n
,
n
a
t
u
r
a
l
n
e
ss
a
n
d
st
r
u
c
t
u
r
e
,
”
I
E
EE
Ac
c
e
ss,
v
o
l
.
6
,
p
p
.
8
0
6
-
8
1
7
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
7
.
2
7
7
6
1
2
6
.
[
8
]
N
.
R
.
S
h
e
n
o
y
a
n
d
A
.
Jat
t
i
,
“
U
l
t
r
a
s
o
u
n
d
i
m
a
g
e
se
g
me
n
t
a
t
i
o
n
t
h
r
o
u
g
h
d
e
e
p
l
e
a
r
n
i
n
g
b
a
s
e
d
i
mp
r
o
v
i
se
d
U
-
N
e
t
,
”
I
n
d
o
n
e
s
i
a
n
J
o
u
r
n
a
l
o
f
E
l
e
c
t
r
i
c
a
l
E
n
g
i
n
e
e
ri
n
g
a
n
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
(
I
J
EE
C
S
)
,
v
o
l
.
2
1
,
n
o
.
3
,
p
p
.
1
4
2
4
-
1
4
3
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
e
c
s.
v
2
1
.
i
3
.
p
p
1
4
2
4
-
1
4
3
4
.
[
9
]
N
.
S
.
Z
o
l
k
i
f
l
i
,
A
.
N
a
z
a
r
i
,
a
n
d
M
.
M
.
M
u
st
a
f
a
,
“
R
e
t
i
n
a
b
l
o
o
d
v
e
sse
l
e
x
t
r
a
c
t
i
o
n
b
a
se
d
o
n
K
i
r
sc
h
's
t
e
m
p
l
a
t
e
me
t
h
o
d
,
”
I
n
d
o
n
e
si
a
n
J
o
u
rn
a
l
o
f
El
e
c
t
r
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
a
n
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
(
I
J
EE
C
S
)
,
v
o
l
.
1
8
,
n
o
.
1
,
p
p
.
3
1
8
-
3
2
5
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
e
c
s.
v
1
8
.
i
1
.
p
p
3
1
8
-
3
2
5
.
[
1
0
]
D
.
D
e
v
a
r
a
j
a
n
d
K
.
S
.
C
.
P
r
a
sa
n
n
a
,
“
D
e
v
e
l
o
p
m
e
n
t
o
f
a
h
y
b
r
i
d
f
r
a
m
e
w
o
r
k
t
o
c
h
a
r
a
c
t
e
r
i
z
e
r
e
d
l
e
si
o
n
s
f
o
r
e
a
r
l
y
d
e
t
e
c
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
,
”
I
n
d
o
n
e
si
a
n
J
o
u
r
n
a
l
o
f
El
e
c
t
r
i
c
a
l
E
n
g
i
n
e
e
ri
n
g
a
n
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
(
I
J
EEC
S
)
,
v
o
l
.
1
3
,
n
o
.
3
,
p
p
.
9
6
2
-
9
7
3
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
e
c
s
.
v
1
3
.
i
3
.
p
p
9
6
2
-
9
7
3
.
[
1
1
]
Z.
X
i
a
o
e
t
a
l
.
,
“
D
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
r
e
t
i
n
a
l
i
m
a
g
e
e
n
h
a
n
c
e
me
n
t
b
a
se
d
o
n
g
a
mm
a
c
o
r
r
e
c
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
Me
d
i
c
a
l
I
m
a
g
i
n
g
a
n
d
H
e
a
l
t
h
I
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
7
,
n
o
.
1
,
p
p
.
1
4
9
-
1
5
4
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
6
6
/
j
mi
h
i
.
2
0
1
7
.
1
9
9
8
.
[
1
2
]
M
.
M
o
u
z
a
i
,
C
.
Ta
r
a
b
e
t
,
a
n
d
A
.
M
u
s
t
a
p
h
a
,
“
L
o
w
-
c
o
n
t
r
a
s
t
X
-
r
a
y
e
n
h
a
n
c
e
me
n
t
u
s
i
n
g
a
f
u
z
z
y
g
a
mm
a
r
e
a
s
o
n
i
n
g
m
o
d
e
l
,
”
Me
d
i
c
a
l
&
Bi
o
l
o
g
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
&
C
o
m
p
u
t
i
n
g
,
v
o
l
.
5
8
,
p
p
.
1
1
7
7
-
1
1
9
7
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
5
1
7
-
020
-
0
2
1
2
2
-
y.
[
1
3
]
K
.
G
.
S
u
ma
a
n
d
V
.
S
.
K
u
m
a
r
,
“
A
q
u
a
n
t
i
t
a
t
i
v
e
a
n
a
l
y
s
i
s
o
f
h
i
s
t
o
g
r
a
m
e
q
u
a
l
i
z
a
t
i
o
n
-
b
a
s
e
d
m
e
t
h
o
d
s
o
n
f
u
n
d
u
s
i
ma
g
e
s
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
,
”
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
B
i
g
D
a
t
a
An
a
l
y
t
i
c
,
2
0
1
9
,
p
p
.
5
5
-
6
3
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
981
-
13
-
0
5
4
4
-
3
_
5
.
[
1
4
]
A
.
M
.
R
e
z
a
,
“
R
e
a
l
i
z
a
t
i
o
n
o
f
t
h
e
c
o
n
t
r
a
st
l
i
mi
t
e
d
a
d
a
p
t
i
v
e
h
i
s
t
o
g
r
a
m
e
q
u
a
l
i
z
a
t
i
o
n
(
C
LA
H
E)
f
o
r
r
e
a
l
-
t
i
m
e
i
m
a
g
e
e
n
h
a
n
c
e
m
e
n
t
,
”
Jo
u
rn
a
l
o
f
VL
S
I
s
i
g
n
a
l
p
ro
c
e
ss
i
n
g
syst
e
m
s
f
o
r
si
g
n
a
l
,
i
m
a
g
e
a
n
d
v
i
d
e
o
t
e
c
h
n
o
l
o
g
y
,
v
o
l
.
3
8
,
n
o
.
1
,
p
p
.
3
5
-
4
4
,
2
0
0
4
,
d
o
i
:
1
0
.
1
0
2
3
/
B
:
V
LSI
.
0
0
0
0
0
2
8
5
3
2
.
5
3
8
9
3
.
8
2
.
[
1
5
]
S
.
K
.
Y
a
d
a
v
,
S
.
K
u
mar
,
B
.
K
u
m
a
r
,
a
n
d
R
.
G
u
p
t
a
,
“
C
o
mp
a
r
a
t
i
v
e
a
n
a
l
y
si
s
o
f
f
u
n
d
u
s
i
ma
g
e
e
n
h
a
n
c
e
me
n
t
i
n
d
e
t
e
c
t
i
o
n
o
f
d
i
a
b
e
t
i
c
R
e
t
i
n
o
p
a
t
h
y
,
”
2
0
1
6
I
EEE
Re
g
i
o
n
1
0
H
u
m
a
n
i
t
a
r
i
a
n
T
e
c
h
n
o
l
o
g
y
C
o
n
f
e
r
e
n
c
e
(
R
1
0
-
H
T
C
)
,
A
g
r
a
,
I
n
d
i
a
,
2
0
1
6
,
p
p
.
1
-
5
,
d
o
i
:
1
0
.
1
1
0
9
/
R
1
0
-
H
TC
.
2
0
1
6
.
7
9
0
6
8
1
4
.
[
1
6
]
F
.
M
.
S
h
a
ms
u
d
e
e
n
a
n
d
G
.
R
a
j
u
,
“
En
h
a
n
c
e
me
n
t
o
f
f
u
n
d
u
s
i
ma
g
e
r
y
,
”
2
0
1
6
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
N
e
x
t
G
e
n
e
ra
t
i
o
n
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
(
I
C
N
G
I
S
)
,
K
o
t
t
a
y
a
m,
I
n
d
i
a
,
2
0
1
6
,
p
p
.
1
-
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
N
G
I
S
.
2
0
1
6
.
7
8
5
4
0
2
2
.
[
1
7
]
M
.
Z
h
o
u
,
K
.
Ji
n
,
S
.
W
a
n
g
,
J
.
Y
e
,
a
n
d
D
.
Q
i
a
n
,
“
C
o
l
o
r
r
e
t
i
n
a
l
i
m
a
g
e
e
n
h
a
n
c
e
me
n
t
b
a
se
d
o
n
l
u
m
i
n
o
si
t
y
a
n
d
c
o
n
t
r
a
st
a
d
j
u
st
me
n
t
,
”
I
EEE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
Bi
o
m
e
d
i
c
a
l
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
6
5
,
n
o
.
3
,
p
p
.
5
2
1
-
5
2
7
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
T
B
M
E.
2
0
1
7
.
2
7
0
0
6
2
7
.
[
1
8
]
S
.
K
u
mar,
S
.
C
h
o
u
d
h
a
r
y
,
R
.
G
u
p
t
a
,
a
n
d
B
.
K
u
mar,
“
P
e
r
f
o
r
ma
n
c
e
e
v
a
l
u
a
t
i
o
n
o
f
j
o
i
n
t
f
i
l
t
e
r
i
n
g
a
n
d
h
i
st
o
g
r
a
m
e
q
u
a
l
i
z
a
t
i
o
n
t
e
c
h
n
i
q
u
e
s
f
o
r
r
e
t
i
n
a
l
f
u
n
d
u
s
i
ma
g
e
e
n
h
a
n
c
e
m
e
n
t
,
”
2
0
1
8
5
t
h
I
EEE
U
t
t
a
r
P
ra
d
e
s
h
S
e
c
t
i
o
n
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
E
l
e
c
t
ri
c
a
l
,
El
e
c
t
r
o
n
i
c
s
a
n
d
C
o
m
p
u
t
e
r
E
n
g
i
n
e
e
r
i
n
g
(
U
PC
O
N
)
,
2
0
1
8
,
p
p
.
1
-
5
,
d
o
i
:
1
0
.
1
1
0
9
/
U
P
C
O
N
.
2
0
1
8
.
8
5
9
7
0
5
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.