I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
43
,
No
.
2
,
A
u
g
u
s
t
2
0
2
6
,
p
p
.
6
2
8
~
6
3
9
I
SS
N:
2
5
0
2
-
4
7
5
2
,
DOI
: 1
0
.
1
1
5
9
1
/ijeecs.v
43
.i
2
.
pp
62
8
-
6
3
9
628
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs.ia
esco
r
e.
co
m
A deep
learning
-
b
a
sed sy
stem f
o
r c
o
ra
l reef
ima
g
e se
g
menta
tion
using
YO
L
O
v
8
w
ith
Ef
ficien
tNe
t
-
B
0
in
Indo
nesia
n w
a
ters
Ra
den Sutia
di
1
,
S
pa
riso
m
a
Viridi
1
,
G
iy
a
nto
2
,
Anda
rt
a
F
.
K
ho
ir
2
,
Rizk
ie
S.
Ut
a
m
a
2
,
E
ls
a
D.
Aulia
2
,
T
ri
A.
H
a
di
2
,
L
ud
i P
a
rwa
da
ni
Aj
i
2,
3
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
a
t
i
o
n
a
l
S
c
i
e
n
c
e
,
F
a
c
u
l
t
y
o
f
M
a
t
h
e
ma
t
i
c
s
a
n
d
N
a
t
u
r
a
l
S
c
i
e
n
c
e
s,
I
n
st
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
B
a
n
d
u
n
g
,
B
a
n
d
u
n
g
,
I
n
d
o
n
e
si
a
2
R
e
s
e
a
r
c
h
C
e
n
t
e
r
f
o
r
O
c
e
a
n
o
g
r
a
p
h
y
,
N
a
t
i
o
n
a
l
R
e
se
a
r
c
h
a
n
d
I
n
n
o
v
a
t
i
o
n
A
g
e
n
c
y
,
I
n
d
o
n
e
si
a
3
M
a
r
i
n
e
E
v
o
l
u
t
i
o
n
a
n
d
Ec
o
l
o
g
y
G
r
o
u
p
,
N
a
t
u
r
a
l
i
s
B
i
o
d
i
v
e
r
si
t
y
C
e
n
t
e
r
,
T
h
e
N
e
t
h
e
r
l
a
n
d
s
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 1
1
,
2
0
2
5
R
ev
is
ed
J
u
n
2
8
,
2
0
2
6
Acc
ep
ted
J
u
l 2
3
,
2
0
2
6
M
a
n
u
a
l
c
o
ra
l
p
o
in
t
c
o
u
n
t
with
e
x
c
e
l
e
x
ten
sio
n
(CP
Ce
)
a
n
a
ly
sis
re
q
u
ires
a
p
p
ro
x
ima
tely
4
–
6
h
o
u
rs
to
p
r
o
c
e
ss
o
n
e
5
0
-
ima
g
e
sta
ti
o
n
,
li
m
it
in
g
th
e
sc
a
le
o
f
c
o
ra
l
re
e
f
m
o
n
i
to
ri
n
g
.
Th
is
st
u
d
y
p
re
se
n
ts
a
n
a
rt
ifi
c
ial
in
tell
ig
e
n
c
e
-
b
a
se
d
wo
rk
fl
o
w
u
sin
g
YO
LOv
8
wi
t
h
a
n
Eff
icie
n
tNe
t
-
B0
b
a
c
k
b
o
n
e
t
o
a
u
to
m
a
te
b
e
n
th
ic
c
o
v
e
r
e
stim
a
ti
o
n
.
A
t
o
t
a
l
o
f
8
,
1
4
7
u
n
d
e
rwa
ter
ima
g
e
s
fro
m
1
5
1
tran
se
c
ts
a
c
ro
ss
3
9
I
n
d
o
n
e
sia
n
c
o
ra
l
re
e
f
sta
ti
o
n
s
we
re
a
n
n
o
tate
d
i
n
to
e
le
v
e
n
b
e
n
th
ic
c
a
teg
o
ries
fo
r
m
o
d
e
l
trai
n
in
g
,
wh
i
le
e
v
a
lu
a
ti
o
n
wa
s
c
o
n
d
u
c
ted
u
sin
g
th
e
2
0
2
1
De
ra
wa
n
Isla
n
d
s
d
a
tas
e
t.
Du
ri
n
g
v
a
li
d
a
ti
o
n
,
YO
LOv
8
a
c
h
iev
e
d
a
n
a
v
e
ra
g
e
m
AP@0
.
5
o
f
0
.
5
8
,
re
c
a
ll
o
f
0
.
7
9
,
a
n
d
a
n
a
v
e
ra
g
e
a
b
so
lu
t
e
p
e
rc
e
n
tag
e
c
o
v
e
r
e
rro
r
o
f
9
.
8
%
c
o
m
p
a
re
d
with
CP
Ce
.
T
h
e
m
o
d
e
l
p
ro
c
e
ss
e
d
e
a
c
h
ima
g
e
in
3
.
1
3
se
c
o
n
d
s,
e
q
u
iv
a
len
t
to
1
5
6
.
4
7
se
c
o
n
d
s
p
e
r
5
0
-
ima
g
e
sta
ti
o
n
,
re
p
re
se
n
ti
n
g
a
9
2
–
1
3
8
×
sp
e
e
d
u
p
o
v
e
r
m
a
n
u
a
l
C
P
Ce
a
n
a
ly
sis.
Th
e
se
re
su
lt
s
s
h
o
w
th
a
t
t
h
e
p
r
o
p
o
se
d
wo
r
k
flo
w
c
a
n
s
u
p
p
o
rt
sc
a
lab
le
a
n
d
n
e
a
r
-
re
a
l
-
ti
m
e
c
o
ra
l
re
e
f
m
o
n
it
o
ri
n
g
a
c
ro
ss
I
n
d
o
n
e
sia
.
K
ey
w
o
r
d
s
:
Ar
tific
ial
in
tellig
en
ce
B
en
th
ic
h
ab
itat c
lass
if
icatio
n
C
o
r
al
E
f
f
icien
tNet
-
B0
I
n
s
tan
ce
s
eg
m
en
tatio
n
YOL
Ov
8
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
:
R
ad
en
Su
tiad
i
Dep
ar
tm
en
t o
f
C
o
m
p
u
tatio
n
al
Scien
ce
,
Facu
lty
o
f
Ma
th
em
ati
cs a
n
d
N
atu
r
al
Scien
ce
s
I
n
s
titu
te
T
ec
h
n
o
lo
g
y
B
an
d
u
n
g
J
l.
Gan
esh
a
n
o
.
1
0
B
an
d
u
n
g
,
W
est
J
av
a
I
n
d
o
n
esia
E
m
ail: su
tiad
ir
ad
en
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
I
n
d
o
n
esia,
th
e
wo
r
l
d
’
s
lar
g
e
s
t
ar
ch
ip
elag
ic
n
atio
n
,
co
m
p
r
is
es
o
v
er
1
7
,
0
0
0
is
lan
d
s
an
d
h
o
s
ts
ap
p
r
o
x
im
ately
5
1
,
0
0
0
k
m
²
o
f
co
r
al
r
ee
f
s
[
1
]
.
T
h
ese
ec
o
s
y
s
tem
s
s
u
p
p
o
r
t
b
io
d
iv
er
s
ity
,
f
is
h
er
ies,
to
u
r
is
m
,
a
n
d
co
astal
p
r
o
tectio
n
[
2
]
,
b
u
t
ar
e
in
cr
ea
s
in
g
ly
th
r
ea
ten
ed
b
y
cl
im
ate
ch
an
g
e
an
d
lo
ca
l
a
n
th
r
o
p
o
g
e
n
ic
p
r
ess
u
r
es
[
3
]
-
[
5
]
.
E
f
f
ec
tiv
e
r
ee
f
co
n
s
er
v
atio
n
th
er
ef
o
r
e
r
e
q
u
ir
es
ac
c
u
r
ate
an
d
s
ca
lab
le
m
o
n
it
o
r
in
g
.
His
to
r
ically
,
co
r
al
r
ee
f
m
o
n
ito
r
i
n
g
in
I
n
d
o
n
esia
r
elied
o
n
co
n
v
e
n
tio
n
al
tr
an
s
ec
t
m
eth
o
d
s
,
s
u
ch
as
p
o
in
t
an
d
lin
e
in
ter
ce
p
t
tr
an
s
ec
ts
,
wh
ich
wer
e
o
f
ten
li
m
ited
to
ac
ce
s
s
ib
le
r
ee
f
s
d
u
e
to
lo
g
is
tical
co
n
s
tr
ain
ts
an
d
r
elian
ce
o
n
SC
UB
A
d
iv
in
g
[
6
]
,
[
7
]
.
C
o
r
al
r
ee
f
m
o
n
ito
r
in
g
h
as
in
cr
ea
s
in
g
ly
ad
o
p
ted
im
ag
e
-
b
ased
m
eth
o
d
s
an
d
s
tan
d
ar
d
ize
d
d
at
a
wo
r
k
f
lo
ws to
im
p
r
o
v
e
d
o
cu
m
en
tatio
n
an
d
co
m
p
a
r
ab
ilit
y
ac
r
o
s
s
m
o
n
ito
r
in
g
p
r
o
g
r
am
s
[
8
]
.
Desp
ite
th
e
b
r
o
ad
e
r
u
s
e
o
f
i
m
ag
e
-
b
ased
r
ee
f
s
u
r
v
ey
s
,
b
e
n
th
ic
class
if
icatio
n
s
till
r
elies
h
ea
v
ily
o
n
tr
ain
ed
ex
p
er
ts
an
d
m
a
n
u
al
an
n
o
tatio
n
to
o
ls
s
u
ch
as C
P
C
e,
wh
ich
ar
e
tim
e
-
co
n
s
u
m
in
g
an
d
s
u
b
ject
to
o
b
s
er
v
er
v
ar
iab
ilit
y
[
9
]
,
[
1
0
]
.
T
h
ese
lim
itatio
n
s
h
av
e
en
co
u
r
a
g
ed
th
e
u
s
e
o
f
ar
tific
ial
in
tellig
en
ce
(
AI
)
to
au
to
m
ate
co
r
al
r
ee
f
im
ag
e
a
n
aly
s
is
,
as
AI
-
b
ased
r
ec
o
g
n
itio
n
s
y
s
tem
s
h
av
e
s
h
o
wn
s
tr
o
n
g
p
er
f
o
r
m
a
n
ce
in
v
ar
io
u
s
im
ag
e
class
if
icatio
n
task
s
[
1
1
]
.
R
ec
e
n
t
ap
p
licatio
n
o
f
YOL
Ov
8
f
o
r
co
r
al
in
s
tan
ce
s
eg
m
en
tatio
n
in
I
n
d
o
n
esian
wate
r
s
ac
h
iev
ed
a
m
AP5
0
o
f
9
8
.
2
%,
p
r
ec
is
io
n
o
f
9
6
.
7
%,
an
d
r
ec
all
o
f
9
5
.
9
%,
in
d
icatin
g
its
p
o
ten
tial
f
o
r
lar
g
e
-
s
ca
le
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
ystem
fo
r
co
r
a
l reef
ima
g
e
s
eg
men
ta
tio
n
u
s
in
g
YOLOv8
w
ith
… (
R
a
d
en
S
u
tia
d
i
)
629
r
ee
f
m
o
n
ito
r
in
g
[
1
2
]
.
I
m
ag
e
s
eg
m
en
tatio
n
is
p
a
r
ticu
lar
ly
r
elev
an
t
b
ec
a
u
s
e
it
id
en
tifie
s
n
o
t
o
n
ly
o
b
ject
class
es
b
u
t
also
th
eir
s
p
atial
ex
ten
t.
Alth
o
u
g
h
s
em
an
tic
s
eg
m
en
tat
io
n
h
as
b
e
en
a
p
p
lied
to
b
e
n
th
ic
h
ab
itat
m
ap
p
in
g
with
m
o
r
e
th
an
8
5
%
o
v
er
all
ac
cu
r
ac
y
u
s
in
g
u
n
s
u
p
er
v
is
ed
alg
o
r
ith
m
s
,
it
d
o
es
n
o
t
p
r
o
v
i
d
e
th
e
o
b
ject
-
lev
el
d
is
cr
im
in
atio
n
o
f
f
er
ed
b
y
in
s
t
an
ce
s
eg
m
en
tatio
n
[
1
3
]
.
YOL
Ov
8
was selec
ted
b
ec
au
s
e
it su
p
p
o
r
ts
b
o
th
o
b
ject
d
etec
t
io
n
an
d
in
s
tan
ce
s
eg
m
en
tatio
n
,
en
ab
lin
g
th
e
id
en
tific
atio
n
o
f
in
d
iv
id
u
al
b
e
n
th
ic
o
b
jects
with
p
ix
el
-
lev
el
m
ask
s
.
T
h
e
d
ef
au
lt
b
ac
k
b
o
n
e
was
r
ep
lace
d
with
E
f
f
icien
tNet
-
B
0
to
im
p
r
o
v
e
co
m
p
u
tatio
n
al
ef
f
icien
cy
,
as
E
f
f
icien
tNet
u
s
es
co
m
p
o
u
n
d
s
ca
lin
g
to
b
alan
ce
n
etwo
r
k
d
ep
th
,
wid
th
,
an
d
in
p
u
t
r
es
o
lu
tio
n
[
1
4
]
.
T
r
ain
in
g
im
ag
es
wer
e
an
n
o
tated
u
s
in
g
R
o
b
o
f
lo
w
to
en
s
u
r
e
lab
elin
g
co
n
s
is
ten
cy
,
an
d
th
e
r
esu
ltin
g
YOL
Ov
8
-
E
f
f
icien
tNe
t
-
B
0
m
o
d
el
was
ev
alu
ated
f
o
r
m
u
lticlas
s
b
en
th
ic
s
eg
m
en
tatio
n
.
T
h
is
s
tu
d
y
co
m
p
ar
ed
th
e
m
o
d
el
-
g
en
e
r
ated
b
en
th
ic
co
v
er
esti
m
ates
wi
th
m
an
u
al
co
r
al
p
o
i
n
t
co
u
n
t
with
ex
ce
l
ex
ten
s
io
n
(
C
P
C
e)
an
aly
s
is
u
s
in
g
th
e
2
0
2
1
Der
awa
n
I
s
lan
d
s
d
ataset
a
s
a
ca
s
e
s
tu
d
y
.
T
h
e
p
r
o
p
o
s
ed
wo
r
k
f
lo
w
aim
s
to
s
u
p
p
o
r
t
s
ca
lab
le
co
r
al
r
ee
f
im
ag
e
an
aly
s
is
f
o
r
m
o
n
ito
r
in
g
ap
p
licatio
n
s
in
I
n
d
o
n
esia.
T
h
is
s
tu
d
y
aim
ed
to
esti
m
ate
b
en
th
ic
co
v
er
u
s
in
g
YOL
Ov
8
-
E
f
f
icien
tNet
-
B
0
,
co
m
p
ar
e
th
e
r
esu
lts
with
m
an
u
al
C
PC
e
an
aly
s
is
,
a
n
d
ev
alu
ate
i
n
f
er
en
ce
s
p
ee
d
f
o
r
s
ca
lab
le
co
r
al
r
ee
f
m
o
n
ito
r
in
g
.
2.
MET
H
O
D
Ph
o
to
g
r
ap
h
ic
d
ata
wer
e
c
o
llected
b
y
th
e
C
o
r
al
R
ee
f
Mo
n
ito
r
in
g
R
esear
ch
T
ea
m
at
3
9
s
u
r
v
ey
s
ites
u
n
d
er
th
e
C
OR
E
MA
P
-
C
T
I
p
r
o
g
r
am
a
n
d
a
r
ch
iv
e
d
in
t
h
e
Na
tio
n
al
Scien
tific
R
ep
o
s
ito
r
y
(
R
I
N)
o
f
B
R
I
N.
T
h
e
d
ataset
co
v
er
s
2
0
1
3
–
2
0
2
1
an
d
in
clu
d
es
8
,
1
4
7
u
n
d
er
wate
r
im
ag
es
f
r
o
m
1
5
1
tr
an
s
ec
ts
,
c
o
n
s
is
tin
g
o
f
7
,
5
5
0
im
ag
es
with
q
u
ad
r
at
f
r
am
es
an
d
5
9
7
im
a
g
es
with
o
u
t
f
r
am
es.
Fo
r
m
o
d
el
ev
alu
atio
n
,
a
s
u
b
s
et
f
r
o
m
th
e
Der
awa
n
I
s
lan
d
s
in
2
0
2
1
was
u
s
ed
b
ec
au
s
e
it
co
n
tain
ed
clea
r
b
en
th
ic
ca
teg
o
r
ies
an
d
h
i
g
h
-
q
u
ality
an
n
o
tated
im
ag
es.
T
h
i
s
s
u
b
s
et
was
ch
o
s
en
d
u
e
to
its
clea
r
r
e
p
r
esen
tatio
n
o
f
v
a
r
io
u
s
b
e
n
th
ic
ca
teg
o
r
ie
s
an
d
its
ec
o
lo
g
ical
im
p
o
r
tan
ce
with
in
th
e
I
n
d
o
n
esian
co
r
al
r
ee
f
n
etwo
r
k
.
T
h
e
u
s
e
o
f
Der
awa
n
as
a
ca
s
e
s
tu
d
y
p
r
o
v
id
es
a
p
r
ac
tical
d
em
o
n
s
tr
atio
n
o
f
th
e
m
o
d
el'
s
p
er
f
o
r
m
an
ce
in
a
r
ea
l
-
wo
r
l
d
s
ettin
g
.
T
h
e
u
n
d
er
wate
r
p
h
o
to
t
r
an
s
ec
t
(
UPT)
tech
n
iq
u
e
was
u
s
ed
t
o
co
llect
c
o
r
al
r
ee
f
b
o
tto
m
im
ag
er
y
.
T
h
is
m
eth
o
d
i
n
v
o
lv
es
estab
lis
h
in
g
a
5
0
-
m
ete
r
tr
an
s
ec
t
lin
e
al
o
n
g
th
e
r
ee
f
s
lo
p
e
at
a
d
ep
th
o
f
ap
p
r
o
x
im
ately
4
–
7
m
eter
s
,
with
p
h
o
to
g
r
ap
h
s
ta
k
en
at
1
-
m
eter
in
ter
v
als
u
s
in
g
4
4
×
5
8
cm
f
r
am
es.
E
ac
h
tr
an
s
ec
t
p
r
o
d
u
ce
d
ap
p
r
o
x
im
ately
5
0
p
h
o
to
f
r
am
es,
tak
en
f
r
o
m
a
d
is
tan
ce
o
f
ap
p
r
o
x
im
ately
1
5
0
c
m
to
en
s
u
r
e
o
p
tim
al
im
ag
e
r
eso
lu
tio
n
an
d
clar
ity
.
T
h
is
ap
p
r
o
ac
h
h
as
b
ee
n
v
alid
ated
as
an
ef
f
ec
tiv
e
m
et
h
o
d
f
o
r
c
o
r
al
r
ee
f
ass
ess
m
en
t
in
r
ec
en
t stu
d
ies
as sh
o
wn
in
Fig
u
r
e
1
[
1
5
]
.
Fig
u
r
e
1
.
Su
r
v
ey
m
a
p
ac
r
o
s
s
th
e
I
n
d
o
n
esian
ar
c
h
ip
elag
o
s
h
o
win
g
d
ata
co
llectio
n
p
o
in
ts
(
in
y
ello
w)
u
s
ed
as
tr
ain
in
g
d
ata
f
o
r
th
e
d
ee
p
lea
r
n
in
g
m
o
d
el
(
YOL
Ov
8
)
in
b
e
n
t
h
ic
h
ab
itat id
en
tific
atio
n
3.
M
AT
E
R
I
AL
B
e
n
t
h
i
c
a
n
d
s
u
b
s
t
r
at
e
c
at
e
g
o
r
ie
s
w
e
r
e
i
d
e
n
t
i
f
ie
d
b
a
s
e
d
o
n
e
s
t
a
b
l
is
h
e
d
t
r
o
p
i
c
a
l
m
a
r
i
n
e
r
e
s
o
u
r
c
e
s
u
r
v
e
y
g
u
i
d
e
l
i
n
e
s
[
1
6
]
a
n
d
s
i
m
p
l
i
f
i
e
d
i
n
t
o
1
1
c
l
a
s
s
es
:
h
a
r
d
c
o
r
a
l
,
d
e
ad
c
o
r
a
l
,
d
e
a
d
c
o
r
a
l
w
i
t
h
a
l
g
a
e
,
s
o
f
t
c
o
r
a
l
,
s
p
o
n
g
e
s
,
m
a
c
r
o
a
l
g
a
e
,
o
t
h
e
r
b
i
o
t
a
,
r
u
b
b
l
e
,
s
a
n
d
,
s
i
lt
,
a
n
d
r
o
c
k
.
T
h
e
d
a
t
a
s
e
t
i
n
cl
u
d
e
d
i
n
f
o
r
m
a
t
i
o
n
o
n
l
o
c
a
t
i
o
n
,
r
e
e
f
c
o
n
d
i
t
i
o
n
,
h
a
b
i
t
a
t
,
d
e
p
t
h
,
a
n
d
lig
h
t
i
n
g
.
A
l
l
f
r
a
m
e
d
i
m
a
g
es
u
s
ed
a
5
8
×
4
4
c
m
s
t
e
e
l
q
u
a
d
r
a
t
as
t
h
e
a
r
e
a
o
f
i
n
t
e
r
es
t
(
A
O
I
)
.
O
v
e
r
al
l
,
8
,
1
4
7
i
m
a
g
e
s
w
i
t
h
a
m
e
d
i
a
n
s
i
z
e
o
f
4
0
0
0
×
3
0
0
0
p
i
x
e
l
s
p
r
o
d
u
c
e
d
2
3
1
,
9
0
6
a
n
n
o
t
a
t
i
o
n
s
,
a
s
s
h
o
w
n
i
n
F
i
g
u
r
e
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
43
,
No
.
2
,
Au
g
u
s
t
20
26
:
628
-
6
3
9
630
Fig
u
r
e
2
.
Dis
tr
ib
u
tio
n
o
f
im
a
g
e
s
izes a
n
d
asp
ec
t r
atio
s
in
th
e
d
ataset
Data
s
et
p
r
ep
ar
atio
n
,
in
clu
d
in
g
an
n
o
tatio
n
,
a
u
g
m
en
tatio
n
,
an
d
d
ata
s
p
litt
in
g
,
was
p
er
f
o
r
m
e
d
u
s
in
g
th
e
clo
u
d
-
b
ased
R
o
b
o
f
lo
w
p
latf
o
r
m
.
I
m
a
g
es
wer
e
lab
eled
in
to
p
r
ed
ef
in
ed
b
e
n
th
ic
ca
teg
o
r
ies,
wh
ile
au
g
m
e
n
tatio
n
s
tep
s
s
u
ch
as
f
lip
p
in
g
,
c
r
o
p
p
in
g
,
a
f
f
in
e
tr
an
s
f
o
r
m
atio
n
,
b
lu
r
r
in
g
,
a
n
d
n
o
is
e
a
d
d
itio
n
wer
e
ap
p
lied
to
im
p
r
o
v
e
tr
ain
in
g
v
ar
iab
ilit
y
[
1
7
].
Simi
lar
co
r
al
s
eg
m
en
tatio
n
s
tu
d
ies
h
av
e
also
u
s
ed
R
o
b
o
f
lo
w
f
o
r
im
ag
e
an
n
o
tatio
n
an
d
ap
p
lied
a
u
g
m
e
n
tatio
n
tec
h
n
iq
u
es
s
u
ch
as
f
lip
p
in
g
,
cr
o
p
p
in
g
,
b
r
ig
h
tn
ess
ad
ju
s
tm
en
t,
a
n
d
n
o
is
e
ad
d
itio
n
t
o
im
p
r
o
v
e
d
ataset
r
o
b
u
s
tn
ess
[
1
8
]
.
T
h
is
wo
r
k
f
lo
w
p
r
o
d
u
ce
d
a
r
ep
r
o
d
u
ci
b
le
d
ataset
f
o
r
YOL
Ov
8
i
n
s
tan
ce
s
eg
m
en
tatio
n
tr
ain
in
g
as sh
o
w
n
in
Fig
u
r
e
3
.
Data
s
ep
ar
atio
n
in
t
o
tr
ain
in
g
,
v
alid
atio
n
,
a
n
d
test
s
ets
was
p
er
f
o
r
m
ed
d
ir
ec
tly
with
in
th
e
R
o
b
o
f
lo
w
p
latf
o
r
m
.
As
s
h
o
wn
i
n
Fig
u
r
e
4
,
an
ex
am
p
le
o
f
co
r
al
o
b
ject
s
eg
m
en
tatio
n
an
n
o
tatio
n
in
a
4
4
×
5
8
cm
²
f
r
am
e
is
d
is
p
lay
ed
in
th
e
R
o
b
o
f
lo
w
in
ter
f
ac
e,
illu
s
tr
atin
g
h
o
w
b
en
th
ic
f
ea
t
u
r
es
ar
e
ap
p
r
o
p
r
iat
ely
lab
eled
.
T
h
ese
s
tep
s
en
s
u
r
e
a
co
m
p
lete
s
tr
u
ctu
r
ed
d
ataset
th
at
is
im
m
ed
iately
r
ea
d
y
to
tr
ai
n
th
e
YOL
Ov
8
in
s
tan
ce
s
eg
m
en
tatio
n
m
o
d
el
.
Fig
u
r
e
3
.
Sh
o
ws
th
e
in
ter
f
ac
e
o
f
th
e
clo
u
d
-
b
ased
ap
p
licatio
n
R
o
b
o
f
lo
w
au
g
m
en
tatio
n
p
r
o
c
ess
Fig
u
r
e
4
.
I
n
s
tan
t seg
m
en
tatio
n
o
f
b
e
n
th
ic
o
b
je
cts u
s
in
g
t
h
e
r
o
b
o
f
lo
w
p
latf
o
r
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
ystem
fo
r
co
r
a
l reef
ima
g
e
s
eg
men
ta
tio
n
u
s
in
g
YOLOv8
w
ith
… (
R
a
d
en
S
u
tia
d
i
)
631
4.
YO
L
O
v
8
B
A
SE
D
SE
G
ME
N
T
A
T
IO
N
MET
H
O
D
I
n
lin
e
with
th
e
e
v
o
lu
tio
n
o
f
t
h
e
YOL
O
s
er
ies
s
tar
tin
g
f
r
o
m
YOL
Ov
1
[
19
]
,
YOL
Ov
2
[2
0
]
,
YOL
Ov
4
[2
1
]
,
YOL
Ov
5
[
2
2
]
,
a
n
d
Y
OL
Ov
8
[
2
3
]
,
th
is
s
tu
d
y
s
ele
cted
th
e
YOL
Ov
8
m
v
ar
ian
t
b
ec
au
s
e
it
o
f
f
er
s
an
o
p
tim
al
b
alan
ce
b
etwe
en
ac
c
u
r
ac
y
a
n
d
co
m
p
u
tatio
n
al
ef
f
i
cien
cy
as
well
as
b
u
ilt
-
in
s
u
p
p
o
r
t
f
o
r
in
s
tan
ce
s
eg
m
en
tatio
n
wh
ich
is
cr
iti
ca
l
in
s
ea
b
ed
co
v
er
a
n
aly
s
is
.
Me
an
wh
ile,
E
f
f
icien
tNet
[
1
4
]
i
n
t
r
o
d
u
ce
d
a
co
m
p
o
u
n
d
-
s
ca
lin
g
a
p
p
r
o
ac
h
t
h
at
b
alan
ce
s
n
etwo
r
k
d
e
p
th
,
wid
th
,
an
d
in
p
u
t
r
eso
lu
tio
n
to
m
ax
im
ize
ac
cu
r
ac
y
p
er
FLOPS.
E
f
f
icien
tNet
-
B
0
was
in
teg
r
ated
as
th
e
YOL
Ov
8
m
b
ac
k
b
o
n
e
to
im
p
r
o
v
e
s
eg
m
en
tatio
n
ef
f
icien
cy
wh
ile
m
ain
tain
in
g
co
m
p
u
tatio
n
al
p
er
f
o
r
m
an
ce
.
T
h
e
r
esu
ltin
g
m
o
d
el
c
o
m
p
r
is
es:
-
B
a
c
k
b
o
n
e
:
E
f
f
i
c
i
e
n
t
N
e
t
-
B
0
(
5
.
3
m
i
l
l
i
o
n
p
a
r
a
m
e
t
e
r
s
;
0
.
3
9
b
i
l
l
i
o
n
F
L
O
P
S
)
f
o
r
l
i
g
h
t
w
e
i
g
h
t
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
.
-
N
e
c
k
:
p
a
t
h
a
g
g
r
e
g
a
t
i
o
n
n
e
t
w
o
r
k
(
P
A
N
e
t
)
f
o
r
m
u
l
t
i
s
c
a
l
e
f
e
a
t
u
r
e
f
u
s
i
o
n
.
-
H
e
a
d
:
A
d
e
c
o
u
p
l
e
d
d
e
t
e
c
t
i
o
n
h
e
a
d
t
h
a
t
h
a
n
d
l
e
s
c
l
a
s
s
i
f
i
c
a
t
i
o
n
,
b
o
u
n
d
i
n
g
b
o
x
r
e
g
r
e
s
s
i
o
n
,
a
n
d
m
a
s
k
p
r
e
d
i
c
t
i
o
n
.
T
h
e
m
o
d
el
was
im
p
lem
e
n
ted
i
n
Py
th
o
n
u
s
in
g
th
e
Ultr
aly
tics
YOL
Ov
8
f
r
am
ewo
r
k
an
d
tr
ai
n
ed
o
n
a
n
NVI
DI
A
A1
0
0
GPU
f
o
r
1
0
0
ep
o
ch
s
with
6
4
0
×
6
4
0
p
x
in
p
u
t
im
ag
es
as
s
h
o
wn
in
Fig
u
r
e
5
.
T
r
ain
in
g
p
ar
am
eter
s
an
d
d
ata
au
g
m
en
t
atio
n
wer
e
ap
p
lied
to
im
p
r
o
v
e
s
tab
ilit
y
,
r
ed
u
ce
o
v
er
f
itti
n
g
,
an
d
in
c
r
ea
s
e
m
o
d
el
r
o
b
u
s
tn
ess
[2
4
]
.
T
h
e
s
y
s
tem
a
ls
o
au
to
m
atica
lly
d
etec
ted
th
e
5
8
×
4
4
c
m
²
f
r
am
e
as
t
h
e
A
OI
an
d
cr
o
p
p
ed
th
is
r
eg
io
n
u
s
in
g
a
cu
s
to
m
Py
th
o
n
s
cr
ip
t
b
ef
o
r
e
s
eg
m
e
n
tatio
n
.
Fi
g
u
r
e
s
5
(
a)
-
5
(
c)
s
u
m
m
ar
izes
th
e
im
ag
e
-
p
r
o
ce
s
s
in
g
wo
r
k
f
lo
w,
in
cl
u
d
in
g
o
r
ig
i
n
al
im
ag
e
in
p
u
t,
AOI
cr
o
p
p
in
g
,
an
d
f
in
al
YOL
Ov
8
-
b
ased
b
en
t
h
ic
s
eg
m
en
tatio
n
.
T
h
e
YOL
Ov
8
ar
c
h
itectu
r
e
is
m
o
d
if
ied
b
y
r
ep
lacin
g
t
h
e
d
ef
au
lt
b
ac
k
b
o
n
e
with
E
f
f
ici
en
tNet
-
B
0
,
a
lig
h
tweig
h
t
co
n
v
o
lu
tio
n
al
n
etwo
r
k
k
n
o
wn
f
o
r
its
p
er
f
o
r
m
an
ce
in
b
alan
cin
g
ef
f
icie
n
cy
an
d
ac
cu
r
ac
y
.
R
elate
d
wo
r
k
o
n
co
r
al
in
s
tan
ce
s
eg
m
en
tatio
n
h
as
s
h
o
wn
th
at
cu
s
to
m
d
o
wn
-
an
d
u
p
-
s
am
p
lin
g
m
o
d
u
les
(
e.
g
.
ADo
wn
_
HW
D)
an
d
f
u
s
io
n
atten
tio
n
m
ec
h
a
n
is
m
s
ca
n
im
p
r
o
v
e
m
AP5
0
b
y
u
p
to
1
0
.
5
%
o
v
er
th
e
b
aselin
e
YOL
Ov
8
n
[
1
8
]
.
T
h
is
m
o
d
if
ic
atio
n
is
im
p
lem
en
ted
b
y
d
ef
i
n
in
g
a
cu
s
to
m
class
n
am
ed
YO
L
Ov
8
E
f
f
icien
tNet,
wh
er
e
th
e
o
u
tp
u
t
o
f
E
f
f
icie
n
tNet
is
p
ass
ed
th
r
o
u
g
h
a
cu
s
to
m
n
ec
k
lay
e
r
(
1
2
8
0
→
5
1
2
co
n
v
o
lu
tio
n
s
)
,
an
d
f
u
r
th
er
to
th
e
h
ea
d
lay
er
f
o
r
b
o
u
n
d
in
g
b
o
x
,
class
,
an
d
m
ask
p
r
ed
ictio
n
.
T
h
e
m
o
d
el
co
n
f
ig
u
r
atio
n
is
m
an
ag
ed
th
r
o
u
g
h
a
YAM
L
f
ile,
wh
ich
d
etails
im
p
o
r
tan
t
co
m
p
o
n
e
n
ts
s
u
ch
as
th
e
b
ac
k
b
o
n
e,
n
ec
k
,
h
ea
d
,
in
p
u
t
im
ag
e
s
ize,
an
d
n
u
m
b
er
o
f
cl
ass
es.
Fo
r
tr
ain
in
g
,
th
e
s
tan
d
ar
d
YOL
Ov
8
p
ip
elin
e
is
u
s
e
d
,
with
o
n
l
y
a
f
ew
ad
ju
s
tm
en
ts
to
s
u
it
th
e
s
p
ec
if
i
c
ar
ch
itectu
r
e
.
T
h
ese
s
ettin
g
s
ar
e
s
p
ec
if
ically
ai
m
ed
at
im
p
r
o
v
in
g
th
e
m
o
d
el’
s
ab
ilit
y
to
ex
tr
ac
t
f
ea
t
u
r
es,
m
ak
in
g
it
m
o
r
e
ef
f
ec
tiv
e
at
an
aly
zin
g
h
ig
h
-
r
eso
lu
tio
n
u
n
d
er
wate
r
im
a
g
es
an
d
ac
cu
r
ately
d
etec
tin
g
s
m
all
o
r
o
v
er
lap
p
in
g
s
ea
b
e
d
o
b
jects.
Fig
u
r
e
6
illu
s
tr
ates
th
e
m
o
d
if
ied
YOL
Ov
8
-
E
f
f
icien
tNet
-
B
0
ar
ch
itectu
r
e,
h
ig
h
lig
h
tin
g
t
h
e
b
ac
k
b
o
n
e
r
ep
lace
m
en
t
u
s
ed
f
o
r
lig
h
tweig
h
t
f
ea
tu
r
e
ex
tr
ac
tio
n
.
T
h
e
ev
alu
ati
o
n
r
esu
lts
s
h
o
wed
th
at
th
e
m
o
d
if
ied
m
o
d
el
ac
h
ie
v
ed
s
tab
le
tr
ain
i
n
g
,
s
tr
o
n
g
g
en
er
aliza
tio
n
ca
p
ab
ilit
y
,
an
d
co
m
p
etitiv
e
s
eg
m
en
tatio
n
p
er
f
o
r
m
a
n
ce
co
m
p
ar
ed
t
o
th
e
b
aselin
e
YOL
Ov
8
m
o
d
el.
T
o
an
aly
ze
t
h
e
p
er
ce
n
t
co
v
er
o
f
b
en
th
ic
a
n
d
s
u
b
s
tr
ate
ca
teg
o
r
ies,
th
e
o
p
en
-
s
o
u
r
ce
s
o
f
twar
e
C
PC
e
v
er
s
io
n
4
.
1
was
u
s
ed
,
ap
p
ly
i
n
g
3
0
r
an
d
o
m
p
o
i
n
ts
to
ea
c
h
im
ag
e
[
2
5
]
.
E
ac
h
p
o
in
t
was
lin
k
ed
to
a
b
en
th
ic
co
m
p
o
n
en
t
o
r
s
u
b
s
tr
ate
with
i
n
th
e
im
a
g
e
f
r
a
m
e
an
d
m
an
u
a
lly
id
en
tifie
d
b
y
a
n
o
b
s
er
v
er
b
ased
o
n
p
r
ed
e
f
in
ed
ca
teg
o
r
ies.
T
h
e
p
er
ce
n
t
co
v
er
f
o
r
ea
ch
ca
teg
o
r
y
was
th
en
ca
lcu
lated
u
s
in
g
th
e
f
o
llo
win
g
f
o
r
m
u
la.
Fig
u
r
e
7
s
h
o
ws th
e
C
PC
e
in
ter
f
ac
e
u
s
ed
to
ass
ig
n
3
0
r
an
d
o
m
p
o
in
ts
f
o
r
m
an
u
al
b
en
th
ic
c
o
v
er
esti
m
atio
n
.
(
a)
(
b
)
(
c)
Fig
u
r
e
5
.
Seq
u
en
tial Y
OL
Ov
8
im
ag
e
-
p
r
o
ce
s
s
in
g
wo
r
k
f
l
o
w:
(
a)
o
r
ig
i
n
al
im
ag
e
with
q
u
ad
r
a
t f
r
am
e,
(
b
)
cr
o
p
p
e
d
AOI
,
an
d
(
c)
s
eg
m
en
tatio
n
r
esu
lt with
b
e
n
t
h
ic
m
ask
s
an
d
co
n
f
id
en
ce
s
co
r
es
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
43
,
No
.
2
,
Au
g
u
s
t
20
26
:
628
-
6
3
9
632
Fig
u
r
e
6
.
Mo
d
if
ied
YOL
Ov
8
ar
ch
itectu
r
e
s
h
o
win
g
th
e
E
f
f
icien
tNet
-
B
0
b
ac
k
b
o
n
e
r
ep
lace
m
en
t
f
o
r
f
ea
tu
r
e
ex
tr
ac
tio
n
(
%
)
=
×
100
Fig
u
r
e
7
.
C
PC
e
in
ter
f
ac
e
s
h
o
win
g
3
0
r
an
d
o
m
p
o
in
ts
u
s
ed
f
o
r
m
an
u
al
b
en
th
ic
c
o
v
er
an
d
s
u
b
s
tr
ate
class
if
i
ca
tio
n
with
in
th
e
q
u
ad
r
at
f
r
am
e
5.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
o
ev
alu
ate
b
o
u
n
d
in
g
b
o
x
d
etec
tio
n
an
d
i
n
s
tan
ce
s
eg
m
en
tatio
n
p
er
f
o
r
m
a
n
ce
,
th
e
m
o
d
el
was
ass
es
s
ed
u
s
in
g
p
r
ec
is
io
n
,
r
ec
a
ll,
F1
-
s
co
r
e,
an
d
m
ea
n
a
v
er
ag
e
p
r
ec
is
io
n
(
m
AP)
at
I
o
U
th
r
esh
o
ld
s
o
f
0
.
5
a
n
d
0
.
5
:0
.
9
5
f
o
r
b
o
th
b
o
u
n
d
in
g
b
o
x
es
an
d
s
eg
m
en
tatio
n
m
ask
s
[
1
2
]
,
[
2
3
]
.
Pre
cisi
o
n
im
p
r
o
v
ed
f
r
o
m
0
.
2
5
t
o
0
.
4
5
,
wh
ile
r
ec
all
in
c
r
ea
s
ed
f
r
o
m
0
.
1
5
t
o
0
.
3
0
.
T
h
e
m
AP5
0
f
o
r
b
o
u
n
d
in
g
b
o
x
es
e
x
ce
e
d
ed
0
.
5
,
wh
ile
t
h
e
s
eg
m
en
tatio
n
m
ask
s
co
r
e
was
s
lig
h
tly
lo
wer
.
I
n
Fig
u
r
e
8
q
u
alitativ
e
an
aly
s
is
d
em
o
n
s
tr
at
ed
th
at
th
e
m
o
d
el
co
n
s
is
ten
tly
id
en
tifie
d
s
m
all
b
en
th
ic
o
b
jects,
alig
n
in
g
with
ea
r
lier
r
esear
ch
th
at
l
ev
er
ag
ed
im
p
r
o
v
ed
YOL
Ov
8
ar
ch
itectu
r
es
u
n
d
er
s
im
ilar
ly
d
em
an
d
in
g
u
n
d
er
wa
ter
co
n
d
itio
n
s
[1
8
]
.
Fig
u
r
e
s
8
(
a)
-
(
d
)
p
r
esen
ts
th
e
tr
ain
in
g
p
er
f
o
r
m
an
ce
cu
r
v
es
f
o
r
b
o
u
n
d
in
g
b
o
x
d
etec
tio
n
an
d
s
eg
m
en
tatio
n
m
ask
s
,
in
cl
u
d
i
n
g
m
AP,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
ystem
fo
r
co
r
a
l reef
ima
g
e
s
eg
men
ta
tio
n
u
s
in
g
YOLOv8
w
ith
… (
R
a
d
en
S
u
tia
d
i
)
633
YOL
Ov
8
s
h
o
wed
s
tr
o
n
g
s
eg
m
en
tatio
n
p
er
f
o
r
m
an
ce
ac
r
o
s
s
m
u
ltip
le
b
en
t
h
ic
ca
teg
o
r
ies,
ec
h
o
in
g
tr
en
d
s
s
ee
n
in
ea
r
lier
c
o
r
al
r
ee
f
im
ag
in
g
s
tu
d
ies
[
1
2
]
,
[
1
3
]
,
[1
8
]
,
wh
ich
r
ep
o
r
te
d
m
AP5
0
s
co
r
es
ab
o
v
e
0
.
9
8
f
o
r
YOL
Ov
8
in
co
r
al
r
ee
f
s
eg
m
e
n
tatio
n
task
s
.
As
s
h
o
wn
in
Fig
u
r
e
9
,
th
e
m
o
d
el
ac
cu
r
ately
d
etec
ted
th
e
s
teel
f
r
am
e
a
r
ea
as
t
h
e
AOI
b
ef
o
r
e
au
to
m
atic
cr
o
p
p
in
g
.
T
h
e
s
eg
m
en
ted
o
u
tp
u
t
id
en
tifie
d
NA
C
as
th
e
d
o
m
in
an
t
class
,
with
co
n
f
id
en
ce
s
co
r
es
r
an
g
i
n
g
f
r
o
m
0
.
3
9
t
o
0
.
8
9
.
T
h
e
m
o
d
el
also
ac
c
u
r
ately
d
e
tecte
d
SP
(
Sp
o
n
g
e)
0
.
4
8
in
m
a
g
en
ta,
o
th
er
b
io
ta
(
OT
)
0
.
4
4
in
p
u
r
p
le,
an
d
d
ea
d
co
r
al
with
alg
ae
(
DC
A)
in
b
r
o
wn
,
with
co
n
f
id
e
n
ce
s
co
r
es
o
f
0
.
3
0
.
T
h
e
s
teel
f
r
a
m
e
its
elf
was
d
etec
ted
in
p
i
n
k
,
wh
ic
h
s
er
v
e
d
as
th
e
p
r
i
m
ar
y
s
eg
m
e
n
tatio
n
b
o
u
n
d
ar
y
as sh
o
w
n
in
Fig
u
r
es
9
(
a)
an
d
(
b
)
.
A
m
AP@
0
.
5
o
f
0
.
7
4
5
was
r
ec
o
r
d
ed
,
co
n
s
is
ten
t
with
b
en
ch
m
ar
k
r
esu
lts
r
ep
o
r
ted
in
s
i
m
ilar
co
r
al
s
eg
m
en
tatio
n
s
tu
d
ies with
in
c
o
m
p
ar
ab
le
u
n
d
e
r
wate
r
en
v
ir
o
n
m
en
ts
[2
6
]
.
A
d
d
itio
n
ally
,
t
h
e
I
o
U
v
alu
es f
o
r
m
o
s
t
class
es
ex
ce
ed
0
.
7
.
T
h
e
p
r
ec
is
io
n
(
0
.
8
2
4
)
an
d
r
ec
all
(
0
.
8
9
3
)
m
etr
ics
f
u
r
th
er
in
d
icate
co
n
s
i
s
ten
t
p
er
f
o
r
m
a
n
ce
,
p
ar
ticu
lar
ly
f
o
r
d
o
m
in
an
t c
ate
g
o
r
ies s
u
ch
as h
ar
d
co
r
als (
HC
)
an
d
s
o
f
t c
o
r
als (
SC
)
.
(
a)
(
b
)
(
c)
(
d
)
Fig
u
r
e
8
.
YOL
Ov
8
tr
ain
in
g
p
e
r
f
o
r
m
a
n
ce
: (
a)
m
AP c
u
r
v
es,
(
b
)
b
o
u
n
d
in
g
b
o
x
p
r
ec
is
io
n
an
d
r
ec
all,
(
c)
m
ask
p
r
ec
is
io
n
an
d
r
ec
all,
an
d
(
d
)
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e
o
v
er
ep
o
ch
s
(
a)
(
b
)
Fig
u
r
e
9
.
YOL
Ov
8
s
eg
m
en
tati
o
n
r
e
s
u
lts
: (
a)
o
r
i
g
in
al
in
p
u
t i
m
ag
e
an
d
(
b
)
im
ag
e
af
ter
m
o
d
el
d
etec
tio
n
s
h
o
win
g
m
u
lti
-
ca
teg
o
r
y
s
eg
m
en
tatio
n
in
u
n
d
er
wate
r
im
a
g
er
y
,
in
clu
d
i
n
g
s
teel
f
r
am
e,
NAC
,
SP
,
DC
A,
an
d
OT
,
p
r
io
r
to
cr
o
p
p
in
g
Qu
alitativ
e
ev
alu
atio
n
r
ev
ea
l
ed
th
at
th
e
s
eg
m
en
tatio
n
r
e
s
u
lts
co
u
ld
b
e
ca
teg
o
r
ize
d
i
n
to
th
r
ee
p
er
f
o
r
m
an
ce
le
v
els,
as
illu
s
tr
ated
in
Fig
u
r
e
1
0
.
I
n
t
h
e
b
es
t
-
ca
s
e
r
esu
lts
,
th
e
m
o
d
el
p
r
o
d
u
ce
d
ac
c
u
r
ate
a
n
d
well
-
d
ef
in
ed
s
eg
m
en
tatio
n
s
f
o
r
m
u
ltip
le
b
en
th
ic
class
es
,
in
clu
d
in
g
AC
(
y
ello
w)
,
SC
(
g
r
ee
n
)
,
NA
C
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
43
,
No
.
2
,
Au
g
u
s
t
20
26
:
628
-
6
3
9
634
(
lig
h
t
g
r
ee
n
)
,
an
d
DC
A
(
b
r
o
w
n
)
as
s
h
o
w
n
in
Fig
u
r
es
1
0
(
a
)
-
10
(
i)
.
Me
d
iu
m
-
q
u
ality
o
u
tp
u
ts
s
h
o
wed
r
ea
s
o
n
ab
ly
ac
cu
r
ate
p
r
ed
ictio
n
s
,
alth
o
u
g
h
s
o
m
e
m
is
class
if
icatio
n
s
o
cc
u
r
r
e
d
-
o
f
te
n
d
u
e
t
o
v
is
u
al
s
im
ilar
ities
b
etwe
en
ca
t
eg
o
r
ies.
I
n
th
e
lo
west
-
q
u
ality
ca
s
es,
d
etec
tio
n
er
r
o
r
s
s
u
ch
as
o
v
e
r
l
ap
p
in
g
m
ask
s
an
d
in
co
r
r
ec
t
class
p
r
ed
ictio
n
s
wer
e
ev
id
en
t,
ty
p
ically
ca
u
s
ed
b
y
p
o
o
r
lig
h
tin
g
c
o
n
d
iti
o
n
s
,
lo
w
co
n
tr
ast,
o
r
am
b
ig
u
o
u
s
b
e
n
th
ic
tex
tu
r
es.
(
a)
(
b
)
(
c)
(
d
)
(
e)
(f)
(
g
)
(
h
)
(
i)
Fig
u
r
e
1
0
.
R
ep
r
esen
tativ
e
s
eg
m
en
tatio
n
r
esu
lts
o
f
th
e
YOL
Ov
8
m
o
d
el
:
(a
)
,
(
b
)
,
(
c)
B
est r
esu
lts
,
(d
)
,
(
e
)
,
(
f
)
m
ed
i
u
m
r
esu
lts
,
an
d
(g
)
,
(
h
)
,
(
i)
r
esu
lts
with
d
etec
tio
n
er
r
o
r
s
An
aly
s
is
o
f
th
e
DE
R
C
0
1
tr
a
n
s
ec
t
d
ata
r
ev
ea
led
an
a
v
er
ag
e
ab
s
o
lu
te
d
i
f
f
er
en
ce
o
f
9
.
7
8
%
b
etwe
en
m
an
u
al
C
PC
e
esti
m
ates
an
d
YOL
Ov
8
p
r
ed
ictio
n
s
.
T
ab
le
1
d
etails
th
e
p
e
r
ce
n
tag
e
co
v
e
r
an
aly
ze
d
b
y
C
PC
e
a
n
d
p
r
e
d
i
c
t
e
d
b
y
Y
O
L
O
v
8
f
o
r
e
a
c
h
b
e
n
t
h
i
c
c
a
t
e
g
o
r
y
,
a
l
o
n
g
w
i
t
h
t
h
e
a
b
s
o
l
u
t
e
e
r
r
o
r
a
n
d
i
n
f
e
r
e
n
c
e
t
i
m
e
.
F
i
g
u
r
e
1
1
co
m
p
ar
es
th
e
p
er
c
en
tag
e
co
v
e
r
esti
m
ates
o
b
tain
ed
f
r
o
m
m
an
u
al
C
PC
e
an
al
y
s
is
an
d
YOL
Ov
8
p
r
ed
ictio
n
f
o
r
th
e
DE
R
C
0
1
tr
an
s
ec
t,
in
clu
d
in
g
t
h
e
o
v
e
r
all
ca
teg
o
r
y
a
v
er
ag
e
.
T
ab
le
1
.
C
o
m
p
a
r
is
o
n
o
f
b
en
th
i
c
co
v
er
p
er
ce
n
tag
es: m
an
u
al
m
eth
o
d
(
C
PC
e)
v
s
AI
(
YOL
Ov
8
)
o
n
tr
an
s
ec
t 0
1
d
er
awa
n
2
0
2
1
d
ata
S
t
a
t
i
o
n
C
a
t
e
g
o
r
y
%
M
a
n
u
a
l
(
C
P
C
e
)
%
A
I
(
Y
O
LO
v
8
)
A
b
so
l
u
t
e
D
i
f
f
e
r
e
n
c
e
(
%)
D
ER
C
0
1
HC
4
7
,
8
0
6
7
,
1
3
1
9
,
3
3
D
ER
C
0
1
D
C
A
2
7
,
8
7
4
,
4
9
2
3
,
3
8
D
ER
C
0
1
SC
1
,
0
0
4
,
4
9
3
,
4
9
D
ER
C
0
1
SP
3
,
1
3
1
1
,
3
5
8
,
2
2
D
ER
C
0
1
OT
0
,
7
3
5
,
0
1
4
,
2
8
D
ER
C
0
1
S
2
6
,
5
6
2
6
,
5
6
0
,
0
0
A
v
e
r
a
g
e
1
7
.
8
5
1
9
.
8
4
9
.
7
8
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
ystem
fo
r
co
r
a
l reef
ima
g
e
s
eg
men
ta
tio
n
u
s
in
g
YOLOv8
w
ith
… (
R
a
d
en
S
u
tia
d
i
)
635
Fig
u
r
e
1
1
.
Per
ce
n
t c
o
v
er
co
m
p
ar
is
o
n
b
e
twee
n
C
PC
e
an
d
Y
OL
Ov
8
at
DE
R
C
0
1
Desp
ite
ce
r
tain
m
is
m
atch
es
s
u
ch
as
th
e
ten
d
en
cy
to
o
v
er
est
im
ate
liv
e
h
ar
d
co
r
al
(
6
7
.
1
%
v
s
.
4
7
.
8
%)
an
d
u
n
d
e
r
esti
m
ate
co
v
er
ed
d
e
ad
co
r
al
with
alg
ae
(
4
.
5
%
v
s
.
2
7
.
9
%),
th
e
m
o
d
el
d
em
o
n
s
tr
ated
a
m
ea
n
ab
s
o
lu
te
d
ev
iatio
n
o
f
9
.
8
%
wh
en
co
m
p
ar
ed
with
C
PC
e
as
s
es
s
m
en
ts
,
in
d
icatin
g
its
r
esu
lts
g
en
er
ally
f
all
with
in
an
a
cc
ep
tab
le
m
ar
g
i
n
o
f
m
an
u
al
i
n
ter
p
r
etatio
n
[
1
3
]
,
[
1
5
]
.
Acr
o
s
s
th
e
ten
Der
awa
n
tr
an
s
ec
t
s
tatio
n
s
,
th
e
liv
e
co
r
al
(
L
C
)
ca
teg
o
r
y
ex
h
ib
ited
a
n
a
v
er
ag
e
a
b
s
o
lu
te
d
if
f
er
en
ce
o
f
1
2
.
8
8
%
b
etwe
en
m
an
u
al
C
PC
e
esti
m
ate
s
an
d
YOL
Ov
8
p
r
ed
ictio
n
s
(
T
ab
le
2
)
.
Fig
u
r
e
1
2
v
is
u
al
izes
th
e
ab
s
o
lu
te
er
r
o
r
f
o
r
ea
ch
b
en
th
ic
ca
teg
o
r
y
ac
r
o
s
s
Der
awa
n
tr
an
s
ec
t
s
tatio
n
s
,
s
h
o
win
g
wh
ich
class
es
co
n
tr
ib
u
ted
m
o
s
t
to
th
e
d
if
f
er
en
ce
b
etwe
en
C
PC
e
an
d
YOL
Ov
8
esti
m
ates.
Fig
u
r
e
1
3
co
m
p
ar
es
b
en
th
ic
co
v
e
r
esti
m
ates
f
r
o
m
C
P
C
e
an
d
YO
L
Ov
8
f
o
r
s
elec
ted
r
ep
r
esen
tativ
e
b
en
t
h
ic
ca
teg
o
r
i
es.
T
ab
le
2
.
L
iv
e
C
o
r
al
(
% L
C
/HC
)
co
m
p
ar
is
o
n
at
Der
awa
n
tr
a
n
s
ec
t statio
n
s
S
t
a
t
i
o
n
%
L
C
(
C
P
C
e
)
%
L
C
(
Y
O
LO
v
8
)
A
b
so
l
u
t
e
D
i
f
f
e
r
e
n
c
e
(
%)
D
ER
C
0
1
4
7
.
8
0
6
7
.
1
3
1
9
.
3
3
D
ER
C
0
2
3
1
.
5
3
3
3
.
3
2
1
.
7
9
D
ER
C
0
3
2
1
.
8
7
5
1
.
0
9
2
9
.
2
2
D
ER
C
0
5
3
9
.
4
7
4
4
.
0
6
4
.
5
9
D
ER
C
0
6
3
7
.
6
9
5
6
.
3
0
1
8
.
6
1
D
ER
C
0
7
2
4
.
1
3
3
9
.
7
4
1
5
.
6
1
D
ER
C
0
8
1
2
.
8
0
2
0
.
8
0
8
.
0
0
D
ER
C
0
9
5
4
.
0
0
4
8
.
9
3
5
.
0
7
D
ER
C
1
0
3
2
.
2
7
4
5
.
6
0
1
3
.
3
3
D
ER
C
1
1
1
9
.
5
3
3
2
.
7
9
1
3
.
2
6
A
v
e
r
a
g
e
1
2
.
8
8
Fig
u
r
e
1
2
.
Ab
s
o
lu
te
er
r
o
r
p
e
r
s
tatio
n
an
d
b
en
t
h
ic
ca
teg
o
r
y
b
e
twee
n
C
PC
e
an
d
YOL
Ov
8
esti
m
ates
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
43
,
No
.
2
,
Au
g
u
s
t
20
26
:
628
-
6
3
9
636
Fig
u
r
e
1
3
.
C
PC
e
an
d
YOL
Ov
8
b
en
th
ic
c
o
v
er
c
o
m
p
a
r
is
o
n
f
o
r
s
el
ec
ted
r
ep
r
esen
tativ
e
ca
teg
o
r
ies
T
h
is
s
tu
d
y
d
em
o
n
s
tr
ates
th
a
t
th
e
YOL
Ov
8
in
s
tan
ce
s
eg
m
en
tatio
n
m
o
d
el,
in
teg
r
ated
with
th
e
E
f
f
icien
tNet
-
B
0
b
ac
k
b
o
n
e,
ef
f
ec
tiv
ely
id
en
tifie
s
elev
en
b
e
n
th
ic
class
es
in
u
n
d
er
wate
r
c
o
r
al
r
ee
f
im
a
g
er
y
.
W
h
en
test
ed
ac
r
o
s
s
ten
tr
an
s
e
ct
s
tatio
n
s
in
th
e
Der
awa
n
I
s
la
n
d
s
,
th
e
m
o
d
el
ac
h
iev
e
d
an
av
er
ag
e
m
AP@
0
.
5
o
f
0
.
5
8
a
n
d
a
r
ec
all
o
f
0
.
7
9
at
th
e
f
in
al
ep
o
ch
.
C
o
m
p
ar
ed
to
m
an
u
al
esti
m
atio
n
s
u
s
in
g
C
PC
e,
th
e
m
o
d
el
s
h
o
wed
a
m
ea
n
ab
s
o
lu
te
d
if
f
e
r
en
ce
in
p
er
ce
n
tag
e
c
o
v
er
o
f
9
.
8
%.
T
h
e
m
o
s
t
s
ig
n
if
ican
t
d
ev
iatio
n
s
wer
e
o
b
s
er
v
ed
in
h
ar
d
co
r
als
(
HC
,
1
9
.
3
%)
an
d
d
ea
d
c
o
r
als
with
al
g
ae
(
DC
A,
2
3
.
4
%),
w
h
ile
s
m
aller
d
if
f
er
e
n
ce
s
wer
e
n
o
ted
in
s
o
f
t c
o
r
als (
SC
,
3
.
5
%),
s
p
o
n
g
e
s
(
SP
,
8
.
2
%),
o
th
er
b
io
ta
(
OT
,
4
.
3
%),
an
d
s
an
d
(
S,
0
.
0
%).
T
h
ese
r
esu
lts
in
d
icate
th
at
th
e
m
o
d
el
is
g
en
er
ally
r
eliab
le,
alth
o
u
g
h
s
ev
er
al
b
en
th
ic
cl
ass
es
s
ti
l
l
r
eq
u
ir
e
f
u
r
th
er
r
ef
in
e
m
en
t.
W
ith
an
av
er
ag
e
in
f
er
en
c
e
tim
e
o
f
3
.
1
3
s
ec
o
n
d
s
p
er
im
ag
e,
o
r
ap
p
r
o
x
im
ately
1
5
6
.
5
s
ec
o
n
d
s
p
er
5
0
-
im
a
g
e
s
tatio
n
,
th
e
m
o
d
el
s
h
o
ws
s
tr
o
n
g
p
o
te
n
tial
f
o
r
n
ea
r
-
r
ea
l
-
tim
e
co
r
al
r
ee
f
m
o
n
ito
r
in
g
.
C
o
n
s
is
ten
t
with
p
r
ev
io
u
s
YOL
O
-
ba
s
ed
co
r
al
m
o
n
ito
r
in
g
s
tu
d
ies
[
2
6
]
,
[
2
7
]
,
th
e
r
esu
lts
s
u
g
g
est
t
h
at
th
e
p
r
o
p
o
s
ed
YOL
Ov
8
-
b
ased
ap
p
r
o
ac
h
is
ef
f
ec
tiv
e
f
o
r
th
e
Der
awa
n
I
s
lan
d
s
d
ataset
an
d
s
ca
lab
le
f
o
r
b
r
o
ad
e
r
co
r
al
r
ee
f
m
o
n
ito
r
in
g
ac
r
o
s
s
I
n
d
o
n
esia.
6.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
p
r
esen
ts
an
ef
f
ici
en
t
an
d
s
ca
lab
le
d
ee
p
lear
n
in
g
-
b
ased
m
eth
o
d
f
o
r
esti
m
atin
g
b
en
th
ic
co
v
er
in
co
r
al
r
ee
f
ec
o
s
y
s
tem
s
.
T
h
e
YOL
Ov
8
-
E
f
f
icien
tNet
-
B
0
m
o
d
el
p
r
o
ce
s
s
ed
im
ag
es
i
n
3
.
1
3
s
p
e
r
im
ag
e,
ac
h
iev
in
g
a
9
2
–
1
3
8
×
s
p
ee
d
u
p
co
m
p
ar
ed
with
m
a
n
u
al
C
PC
e
an
aly
s
is
.
T
h
e
m
o
d
el
s
h
o
we
d
s
tr
o
n
g
s
eg
m
en
tatio
n
p
er
f
o
r
m
an
ce
,
p
ar
ticu
lar
ly
f
o
r
d
o
m
in
an
t
class
es
s
u
ch
as
h
ar
d
co
r
als
(
HC
)
an
d
n
o
n
-
Acr
o
p
o
r
a
co
r
als
(
NAC),
with
I
o
U
v
alu
es
ex
ce
ed
i
n
g
0
.
7
in
m
an
y
ca
s
es.
Ho
wev
er
,
p
er
f
o
r
m
a
n
ce
m
ay
s
till
b
e
af
f
ec
ted
b
y
ch
allen
g
in
g
im
ag
e
co
n
d
itio
n
s
,
in
clu
d
in
g
u
n
ev
en
lig
h
tin
g
an
d
o
v
er
lap
p
in
g
s
m
all
b
en
th
ic
co
lo
n
ies.
Fu
tu
r
e
wo
r
k
s
h
o
u
ld
ex
p
an
d
th
e
tr
ain
in
g
d
ataset
to
b
etter
r
ep
r
esen
t
I
n
d
o
n
esia’
s
d
i
v
er
s
e
r
ee
f
e
n
v
ir
o
n
m
en
ts
,
in
clu
d
in
g
tu
r
b
id
wate
r
s
an
d
d
ee
p
er
r
ee
f
z
o
n
es,
an
d
v
alid
ate
th
e
wo
r
k
f
lo
w
i
n
r
ea
l
-
tim
e
m
o
n
ito
r
i
n
g
s
ce
n
ar
i
o
s
th
r
o
u
g
h
e
d
g
e
-
b
ased
d
ev
ices o
r
in
teg
r
atio
n
with
C
PC
e.
ACK
NO
WL
E
DG
M
E
N
T
S
W
e
wo
u
ld
lik
e
to
ex
p
r
ess
o
u
r
d
ee
p
est
g
r
atitu
d
e
to
th
e
in
d
iv
i
d
u
als
wh
o
h
av
e
s
ig
n
if
ican
tly
ass
is
ted
in
th
e
wr
itin
g
o
f
th
is
p
ap
er
Ka
h
f
i
R
izk
y
K,
an
d
I
r
f
an
Kam
f
o
n
o
.
Valu
ab
le
in
p
u
t a
n
d
tech
n
ical
d
is
cu
s
s
io
n
s
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
was
s
u
p
p
o
r
t
ed
,
in
p
ar
t,
b
y
t
h
e
Au
s
tr
alia
Awa
r
d
s
I
n
d
o
n
esia
th
r
o
u
g
h
th
e
Had
i So
esas
tr
o
Priz
e.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
at
e
co
llab
o
r
atio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
ystem
fo
r
co
r
a
l reef
ima
g
e
s
eg
men
ta
tio
n
u
s
in
g
YOLOv8
w
ith
… (
R
a
d
en
S
u
tia
d
i
)
637
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
R
ad
en
Su
tiad
i
✓
✓
✓
✓
✓
✓
✓
✓
Sp
ar
is
o
m
a
Vir
id
i
✓
✓
✓
✓
Giy
an
to
✓
✓
✓
✓
✓
✓
An
d
ar
ta
F.
Kh
o
ir
✓
✓
R
izk
ie
S.
Utam
a
✓
✓
✓
E
ls
a
D.
Au
lia
✓
✓
T
r
i A
.
Had
i
✓
✓
✓
✓
✓
L
u
d
i Par
wad
an
i A
ji
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
T
h
e
au
th
o
r
s
d
ec
lar
e
th
at
th
e
r
e
is
n
o
co
n
f
lict o
f
in
ter
est r
elate
d
to
th
is
r
esear
ch
.
I
NF
O
RM
E
D
CO
NS
E
N
T
No
t
ap
p
licab
le.
T
h
is
s
tu
d
y
d
id
n
o
t
in
v
o
lv
e
h
u
m
an
p
ar
ticip
an
ts
,
id
en
tifia
b
le
p
er
s
o
n
al
d
ata,
o
r
clin
ical
p
r
o
ce
d
u
r
es r
eq
u
i
r
in
g
in
f
o
r
m
e
d
co
n
s
en
t.
E
T
H
I
CAL AP
P
RO
V
AL
No
t
ap
p
licab
le.
T
h
is
s
tu
d
y
d
i
d
n
o
t
i
n
v
o
lv
e
h
u
m
an
p
a
r
ticip
a
n
ts
o
r
an
im
al
s
u
b
jects
an
d
th
e
r
ef
o
r
e
d
id
n
o
t r
eq
u
ir
e
eth
ical
ap
p
r
o
v
al.
DATA AV
AI
L
AB
I
L
I
T
Y
Der
iv
ed
d
ata
s
u
p
p
o
r
tin
g
th
e
f
i
n
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
av
ailab
le
f
r
o
m
th
e
co
r
r
esp
o
n
d
i
n
g
au
th
o
r
,
R
.
S.,
u
p
o
n
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
L.
B
u
r
k
e
,
E.
S
e
l
i
g
,
a
n
d
M
.
S
p
a
l
d
i
n
g
,
Re
e
f
s
a
t
Ri
s
k
i
n
S
o
u
t
h
e
a
st
Asi
a
.
2
0
0
2
.
[
2
]
H
.
C
e
s
a
r
,
C
.
G
.
L
u
n
d
i
n
,
S
.
B
e
t
t
e
n
c
o
u
r
t
,
a
n
d
J.
D
i
x
o
n
,
“
I
n
d
o
n
e
s
i
a
n
c
o
r
a
l
r
e
e
f
s
-
A
n
e
c
o
n
o
m
i
c
a
n
a
l
y
s
i
s
o
f
a
p
r
e
c
i
o
u
s
b
u
t
t
h
r
e
a
t
e
n
e
d
r
e
so
u
r
c
e
,
”
Am
b
i
o
,
v
o
l
.
2
6
,
n
o
.
6
,
p
p
.
3
4
5
–
3
5
0
,
1
9
9
7
,
d
o
i
:
1
0
.
1
0
1
6
/
S
0
9
6
7
-
0
6
5
3
(
9
8
)
8
0
6
3
7
-
6.
[
3
]
G
.
B
a
u
m,
H
.
I
.
J
a
n
u
a
r
,
S
.
C
.
A
.
F
e
r
se,
a
n
d
A
.
K
u
n
z
m
a
n
n
,
“
Lo
c
a
l
a
n
d
r
e
g
i
o
n
a
l
i
m
p
a
c
t
s
o
f
p
o
l
l
u
t
i
o
n
o
n
c
o
r
a
l
r
e
e
f
s
a
l
o
n
g
t
h
e
t
h
o
u
sa
n
d
i
s
l
a
n
d
s
n
o
r
t
h
o
f
t
h
e
me
g
a
c
i
t
y
J
a
k
a
r
t
a
,
I
n
d
o
n
e
si
a
,
”
PLo
S
O
n
e
,
v
o
l
.
1
0
,
n
o
.
9
,
p
p
.
1
–
2
6
,
2
0
1
5
,
d
o
i
:
1
0
.
1
3
7
1
/
j
o
u
r
n
a
l
.
p
o
n
e
.
0
1
3
8
2
7
1
.
[
4
]
L.
H
.
D
e
C
l
i
p
p
e
l
e
e
t
a
l
.
,
“
E
v
a
l
u
a
t
i
n
g
a
n
n
u
a
l
se
v
e
r
e
c
o
r
a
l
b
l
e
a
c
h
i
n
g
r
i
sk
f
o
r
m
a
r
i
n
e
p
r
o
t
e
c
t
e
d
a
r
e
a
s
a
c
r
o
ss
I
n
d
o
n
e
si
a
,
”
Ma
r
i
n
e
P
o
l
i
c
y
,
v
o
l
.
1
4
8
,
n
o
.
N
o
v
e
m
b
e
r
2
0
2
2
,
p
.
1
0
5
4
2
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
mar
p
o
l
.
2
0
2
2
.
1
0
5
4
2
8
.
[
5
]
S
.
S
u
l
l
y
,
G
.
H
o
d
g
s
o
n
,
a
n
d
R
.
v
a
n
W
o
e
s
i
k
,
“
P
r
e
se
n
t
a
n
d
f
u
t
u
r
e
b
r
i
g
h
t
a
n
d
d
a
r
k
s
p
o
t
s
f
o
r
c
o
r
a
l
r
e
e
f
s
t
h
r
o
u
g
h
c
l
i
m
a
t
e
c
h
a
n
g
e
,
”
G
l
o
b
a
l
C
h
a
n
g
e
B
i
o
l
o
g
y
,
v
o
l
.
2
8
,
n
o
.
1
5
,
p
p
.
4
5
0
9
–
4
5
2
2
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
1
1
/
g
c
b
.
1
6
0
8
3
.
[
6
]
A
.
R
.
H
a
l
f
o
r
d
a
n
d
I
.
U
r
b
i
n
a
-
b
a
r
r
e
t
o
,
“
W
h
i
c
h
m
e
t
h
o
d
f
o
r
w
h
i
c
h
p
u
r
p
o
s
e
?
a
c
o
m
p
a
r
i
s
o
n
o
f
l
i
n
e
i
n
t
e
r
c
e
p
t
t
r
a
n
s
e
c
t
a
n
d
u
n
d
e
r
w
a
t
e
r
p
h
o
t
o
g
r
a
m
m
e
t
r
y
m
e
t
h
o
d
s
f
o
r
c
o
r
a
l
r
e
e
f
s
u
r
v
e
y
s
,
”
F
r
o
n
t
i
e
r
s
i
n
M
a
r
i
n
e
S
c
i
e
n
c
e
,
v
o
l
.
8
,
n
o
.
M
a
y
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
8
9
/
f
m
a
r
s
.
2
0
2
1
.
6
3
6
9
0
2
.
[
7
]
J.
D
.
P
r
a
s
e
t
y
a
a
n
d
D
.
H
.
S
a
n
t
o
s
o
,
“
C
o
m
p
a
r
a
t
i
v
e
st
u
d
y
o
f
p
o
i
n
t
i
n
t
e
r
c
e
p
t
t
r
a
n
sec
t
(
P
I
T)
met
h
o
d
a
n
d
u
n
d
e
r
w
a
t
e
r
p
h
o
t
o
t
r
a
n
s
e
c
t
(
U
P
T)
t
o
c
a
l
c
u
l
a
t
e
h
a
r
d
c
o
r
a
l
c
o
v
e
r
p
e
r
c
e
n
t
a
g
e
,
”
J
u
r
n
a
l
I
l
m
i
a
h
Pe
r
i
k
a
n
a
n
d
a
n
K
e
l
a
u
t
a
n
,
v
o
l
.
1
4
,
n
o
.
2
,
p
p
.
4
0
4
–
4
1
0
,
2
0
2
2
,
d
o
i
:
1
0
.
2
0
4
7
3
/
j
i
p
k
.
v
1
4
i
2
.
3
2
3
9
3
.
[
8
]
N
.
P
r
a
si
l
D
e
l
a
v
a
l
,
J.
W
i
c
q
u
a
r
t
,
F
.
S
t
a
u
b
,
a
n
d
S
.
P
l
a
n
e
s,
S
t
a
t
u
s o
f
c
o
r
a
l
r
e
e
f
m
o
n
i
t
o
ri
n
g
:
An
a
ssessm
e
n
t
o
f
m
e
t
h
o
d
s
a
n
d
d
a
t
a
a
t
t
h
e
n
a
t
i
o
n
a
l
l
e
v
e
l
,
2
0
2
1
.
[
9
]
O
.
B
e
i
j
b
o
m
e
t
a
l
.
,
“
T
o
w
a
r
d
s
a
u
t
o
ma
t
e
d
a
n
n
o
t
a
t
i
o
n
o
f
b
e
n
t
h
i
c
s
u
r
v
e
y
i
ma
g
e
s:
V
a
r
i
a
b
i
l
i
t
y
o
f
h
u
m
a
n
e
x
p
e
r
t
s
a
n
d
o
p
e
r
a
t
i
o
n
a
l
m
o
d
e
s
o
f
a
u
t
o
m
a
t
i
o
n
,
”
PL
o
S
O
n
e
,
v
o
l
.
1
0
,
n
o
.
7
,
p
p
.
1
–
2
2
,
2
0
1
5
,
d
o
i
:
1
0
.
1
3
7
1
/
j
o
u
r
n
a
l
.
p
o
n
e
.
0
1
3
0
3
1
2
.
[
1
0
]
I
.
D
.
W
i
l
l
i
a
ms
e
t
a
l
.
,
“
Le
v
e
r
a
g
i
n
g
a
u
t
o
m
a
t
e
d
i
ma
g
e
a
n
a
l
y
s
i
s
t
o
o
l
s
t
o
t
r
a
n
s
f
o
r
m
o
u
r
c
a
p
a
c
i
t
y
t
o
a
ssess
s
t
a
t
u
s
a
n
d
t
r
e
n
d
s
o
n
c
o
r
a
l
r
e
e
f
s,”
Fr
o
n
t
i
e
rs i
n
M
a
ri
n
e
S
c
i
e
n
c
e
,
v
o
l
.
6
,
n
o
.
A
P
R
,
p
p
.
1
–
1
4
,
2
0
1
9
,
d
o
i
:
1
0
.
3
3
8
9
/
f
mars
.
2
0
1
9
.
0
0
2
2
2
.
[
1
1
]
Y
.
Zh
a
n
g
,
“
M
u
l
t
i
-
f
a
c
t
o
r
s
a
n
a
l
y
si
s
u
s
i
n
g
v
i
s
u
a
l
i
z
a
t
i
o
n
s
a
n
d
S
H
A
P
:
c
o
mp
r
e
h
e
n
s
i
v
e
c
a
se
a
n
a
l
y
s
i
s
o
f
t
e
n
n
i
s
r
e
su
l
t
s
f
o
r
e
c
a
st
i
n
g
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Ad
v
a
n
c
e
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
6
,
n
o
.
1
,
p
p
.
1
4
3
–
1
5
2
,
2
0
2
5
,
d
o
i
:
1
0
.
1
4
5
6
9
/
I
JA
C
S
A
.
2
0
2
5
.
0
1
6
0
1
1
4
.
[
1
2
]
W
.
M
.
H
a
ssa
n
u
d
i
n
,
V
.
G
.
U
t
o
mo
,
a
n
d
R
.
A
p
r
i
y
a
n
t
o
,
“
F
i
n
e
-
g
r
a
i
n
e
d
a
n
a
l
y
si
s
o
f
c
o
r
a
l
i
n
s
t
a
n
c
e
se
g
me
n
t
a
t
i
o
n
u
s
i
n
g
Y
O
LO
v
8
mo
d
e
l
s
,”
S
i
n
k
ro
n
,
v
o
l
.
8
,
n
o
.
2
,
p
p
.
1
0
4
7
–
1
0
5
5
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
3
9
5
/
si
n
k
r
o
n
.
v
8
i
2
.
1
3
5
8
3
.
[
1
3
]
H
.
Zh
a
n
g
,
A
.
G
r
u
e
n
,
a
n
d
M
.
Li
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
sem
a
n
t
i
c
se
g
me
n
t
a
t
i
o
n
o
f
c
o
r
a
l
i
m
a
g
e
s
i
n
u
n
d
e
r
w
a
t
e
r
p
h
o
t
o
g
r
a
m
me
t
r
y
,”
I
S
PR
S
An
n
a
l
s
o
f
t
h
e
P
h
o
t
o
g
ra
m
m
e
t
ry,
Re
m
o
t
e
S
e
n
s
i
n
g
a
n
d
S
p
a
t
i
a
l
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
5
,
n
o
.
2
,
p
p
.
3
4
3
–
3
5
0
,
2
0
2
2
,
d
o
i
:
1
0
.
5
1
9
4
/
i
s
p
r
s
-
a
n
n
a
l
s
-
V
-
2
-
2
0
2
2
-
3
4
3
-
2
0
2
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.