I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
,
p
p
.
329
~
337
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
1
.
p
p
3
2
9
-
3
3
7
329
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Cha
ng
e det
e
ction
a
nd clas
sifica
tion
o
f
sa
tellit
e
ima
g
es using
co
nv
o
lutiona
l neural network
Ra
g
ha
v
endra
Srini
v
a
s
a
ia
h
1
,
Sa
nto
s
h K
um
a
r
J
a
nk
a
t
t
i
2
,
M
a
njuna
t
h Ra
m
a
nn
a
L
a
m
a
ni
3
,
Nira
nja
na
Sh
ra
v
a
na
bela
g
o
l
a
J
ina
cha
nd
ra
4
1
D
e
p
a
r
t
m
e
n
t
o
f
A
r
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
a
n
d
D
a
t
a
S
c
i
e
n
c
e
,
S
c
h
o
o
l
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
C
H
R
I
S
T
U
n
i
v
e
r
si
t
y
,
B
a
n
g
a
l
o
r
e
,
I
n
d
i
a
2
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
T
e
c
h
n
o
l
o
g
y
,
S
c
h
o
o
l
o
f
E
n
g
i
n
e
e
r
i
n
g
,
D
a
y
a
n
a
n
d
a
S
a
g
a
r
U
n
i
v
e
r
s
i
t
y
,
B
a
n
g
a
l
o
r
e
,
I
n
d
i
a
3
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
,
M
o
o
d
l
a
k
a
t
t
e
I
n
st
i
t
u
t
e
o
f
Te
c
h
n
o
l
o
g
y
,
K
u
n
d
a
p
u
r
a
,
I
n
d
i
a
4
D
e
p
a
r
t
me
n
t
o
f
M
e
c
h
a
n
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
S
c
h
o
o
l
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
Te
c
h
n
o
l
o
g
y
,
C
H
R
I
S
T
U
n
i
v
e
r
si
t
y
,
B
a
n
g
a
l
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
No
v
1
,
2
0
2
4
R
ev
is
ed
Oct
1
9
,
2
0
2
5
Acc
ep
ted
J
an
1
0
,
2
0
2
6
S
a
telli
te
a
n
d
a
ir
b
o
r
n
e
ima
g
e
ry
,
c
o
ll
e
c
ti
v
e
l
y
k
n
o
wn
a
s
e
a
rth
o
b
se
rv
a
ti
o
n
ima
g
e
ry
,
a
re
ima
g
e
s
o
f
t
h
e
e
a
rth
c
o
ll
e
c
ted
fro
m
sp
a
c
e
b
o
r
n
e
o
r
a
irb
o
r
n
e
p
latfo
rm
s
su
c
h
a
s
sa
telli
tes
a
n
d
a
i
rc
ra
ft
.
Ov
e
r
t
h
e
las
t
1
0
0
y
e
a
rs,
wi
th
t
h
e
fa
st
d
e
v
e
lo
p
m
e
n
t
o
f
a
v
iatio
n
,
sp
a
c
e
e
x
p
lo
ra
ti
o
n
,
a
n
d
ima
g
in
g
tec
h
n
o
l
o
g
ies
,
th
e
c
o
m
in
g
t
o
g
e
th
e
r
o
f
th
e
se
tec
h
n
o
l
o
g
ies
h
a
s
b
e
e
n
in
e
v
it
a
b
le.
Eart
h
o
b
se
rv
a
ti
o
n
ima
g
e
ry
h
a
s
m
a
n
y
a
p
p
li
c
a
t
io
n
s
i
n
re
g
io
n
a
l
p
la
n
n
i
n
g
,
g
e
o
lo
g
y
,
re
c
o
n
n
a
issa
n
c
e
,
fish
i
n
g
,
m
e
teo
ro
l
o
g
y
,
o
c
e
a
n
o
g
ra
p
h
y
,
a
g
ricu
lt
u
re
,
b
io
d
iv
e
rsit
y
c
o
n
se
rv
a
ti
o
n
,
f
o
re
stry
,
la
n
d
sc
a
p
e
,
i
n
telli
g
e
n
c
e
,
c
a
rto
g
ra
p
h
y
,
e
d
u
c
a
ti
o
n
,
a
n
d
wa
rfa
re
.
Wi
th
th
e
rise
in
t
h
e
n
u
m
b
e
r
o
f
t
h
e
se
a
irb
o
rn
e
a
n
d
sp
a
c
e
b
o
rn
e
ima
g
i
n
g
p
latfo
rm
s
b
e
in
g
d
e
p
lo
y
e
d
b
y
g
o
v
e
rn
m
e
n
t
a
n
d
p
riv
a
te
e
n
ti
ti
e
s
a
li
k
e
,
th
e
c
a
p
a
b
i
li
ty
to
sif
t
th
r
o
u
g
h
a
n
d
a
n
a
ly
z
e
v
a
st
a
m
o
u
n
ts
o
f
d
a
ta
g
e
n
e
ra
ted
b
y
th
e
se
p
latf
o
rm
s
is
t
h
e
n
e
e
d
o
f
th
e
h
o
u
r
.
Wi
t
h
t
h
e
e
x
p
o
n
e
n
ti
a
l
imp
ro
v
e
m
e
n
t
in
t
h
e
c
o
m
p
u
tatio
n
a
l
c
a
p
a
b
il
it
ies
o
f
c
o
m
p
u
ters
o
v
e
r
th
e
las
t
h
a
lf
a
c
e
n
tu
ry
,
a
n
a
ly
sts
a
re
e
x
c
e
e
d
in
g
l
y
m
o
v
in
g
to
wa
rd
s
th
e
p
ra
c
ti
c
e
o
f
a
rti
ficia
l
in
telli
g
e
n
c
e
,
m
a
c
h
in
e
le
a
rn
in
g
(M
L)
,
a
n
d
c
o
m
p
u
ter
v
isio
n
so
l
u
ti
o
n
s
to
a
u
t
o
m
a
te
a
larg
e
p
a
rt
o
f
t
h
e
p
ro
c
e
ss
e
s
e
m
p
lo
y
e
d
i
n
a
n
a
l
y
z
in
g
e
a
rth
o
b
se
rv
a
ti
o
n
ima
g
e
ry
.
T
h
is
wo
r
k
re
c
o
m
m
e
n
d
s
a
wo
rk
fl
o
w
t
o
p
e
rc
e
iv
e
a
n
d
c
las
sify
c
h
a
n
g
e
s
in
e
a
rth
o
b
se
rv
a
t
io
n
ima
g
e
ry
o
f
a
g
i
v
e
n
a
re
a
b
y
u
ti
li
z
in
g
th
e
v
a
st f
lex
ib
il
it
y
t
h
a
t
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
s
(CNN
)
p
ro
v
id
e
.
K
ey
w
o
r
d
s
:
Ar
tific
ial
in
tellig
en
ce
C
o
m
p
u
ter
v
is
io
n
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
Ma
ch
in
e
lear
n
in
g
Satellite im
ag
e
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
ag
h
av
en
d
r
a
Srin
iv
asaiah
Dep
ar
tm
en
t o
f
Ar
tific
ial
I
n
tellig
en
ce
,
Ma
ch
in
e
L
ea
r
n
in
g
an
d
Data
Scien
ce
Sch
o
o
l o
f
E
n
g
in
ee
r
in
g
a
n
d
T
e
ch
n
o
lo
g
y
,
C
HR
I
ST
Un
iv
er
s
ity
B
an
g
alo
r
e,
I
n
d
ia
E
m
ail:
r
ag
h
av
.
tr
g
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
m
o
s
t
ess
e
n
tia
l
t
as
k
is
s
a
v
in
g
l
iv
es
an
d
p
r
o
te
cti
n
g
t
h
e
i
n
f
r
astr
u
ct
u
r
e
f
r
o
m
n
a
tu
r
a
l
d
is
aste
r
s
.
T
h
e
m
o
s
t c
o
m
m
o
n
f
o
r
m
s
o
f
n
a
tu
r
a
l
d
is
aste
r
s
a
r
e
la
n
d
s
li
d
es
a
n
d
e
a
r
t
h
q
u
ak
es.
Sat
ell
ite
i
m
a
g
es
a
r
e
u
s
ed
t
o
p
r
e
d
ic
t t
h
e
ar
ea
s
o
f
ea
r
t
h
q
u
ak
es
a
n
d
la
n
d
s
lid
es
[
1
]
.
A
lt
h
o
u
g
h
th
e
r
e
a
r
e
n
u
m
e
r
o
u
s
u
s
es
o
f
r
em
o
t
e
s
e
n
s
i
n
g
(
R
S
)
s
at
elli
te
im
a
g
e
r
y
,
o
b
tai
n
i
n
g
th
em
is
m
o
r
e
d
if
f
i
cu
lt.
T
h
ese
im
a
g
es a
r
e
u
s
e
d
to
m
i
n
e
th
e
d
ata
p
r
ese
n
t i
n
th
em
.
S
o
,
s
ate
llit
e
im
a
g
es
ar
e
m
o
s
t
i
m
p
o
r
ta
n
t
f
o
r
la
n
d
c
o
v
e
r
a
n
al
y
s
is
a
n
d
a
n
es
s
en
t
ial
f
o
u
n
d
at
io
n
f
o
r
la
n
d
u
s
e
[
2
]
–
[
4
]
.
I
m
ag
es
o
f
th
e
wo
r
ld
tak
e
n
f
r
o
m
s
p
ac
eb
o
r
n
e
o
r
air
b
o
r
n
e
p
latf
o
r
m
s
,
s
u
ch
as
s
atellites
an
d
air
cr
af
t,
ar
e
r
ef
er
r
ed
t
o
as
s
atellite
an
d
air
b
o
r
n
e
im
ag
in
g
,
also
k
n
o
wn
as
ea
r
th
o
b
s
er
v
atio
n
im
ag
er
y
.
W
ith
th
e
u
s
es
o
f
th
ese
im
ag
es
r
an
g
in
g
f
r
o
m
m
eteo
r
o
lo
g
y
to
i
n
tellig
en
ce
an
d
r
ec
o
n
n
aiss
an
c
e,
th
e
s
h
ee
r
n
u
m
b
er
an
d
d
en
s
i
ty
at
wh
ich
p
r
iv
ate
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
329
-
3
3
7
330
an
d
p
u
b
lic
o
r
g
a
n
izatio
n
s
g
e
n
er
ate
th
ese
im
a
g
es
h
av
e
p
o
s
ed
th
e
c
h
allen
g
e
o
f
ef
f
ec
ti
v
ely
an
al
y
zin
g
an
d
ex
tr
ac
tin
g
m
ea
n
i
n
g
f
u
l in
s
ig
h
ts
f
r
o
m
t
h
em
to
cr
ea
te
ac
tio
n
ab
l
e
in
f
o
r
m
atio
n
.
I
d
en
tify
in
g
c
h
an
g
es
o
n
t
h
e
ea
r
th
'
s
s
u
r
f
ac
e
is
th
e
m
o
s
t
cr
u
ci
al
u
s
e
o
f
R
S
[
5
]
.
T
wo
o
r
m
o
r
e
s
atellite
p
h
o
to
s
o
b
tain
e
d
at
v
ar
io
u
s
p
er
io
d
s
ar
e
u
s
u
ally
co
m
p
ar
e
d
as
a
p
ar
t
o
f
th
e
ch
an
g
e
d
etec
tio
n
(
C
D)
an
d
ca
teg
o
r
izatio
n
p
r
o
ce
s
s
.
T
h
e
im
ag
es
ar
e
p
r
e
-
p
r
o
ce
s
s
ed
an
d
th
en
an
aly
ze
d
u
s
in
g
v
ar
io
u
s
tech
n
iq
u
es,
s
u
ch
as
ch
an
g
e
v
ec
to
r
a
n
aly
s
is
,
im
ag
e
d
if
f
er
e
n
cin
g
,
an
d
m
ac
h
in
e
lear
n
in
g
(
ML
)
alg
o
r
ith
m
s
.
T
h
e
r
esu
ltan
t
ca
teg
o
r
izatio
n
an
d
ch
an
g
e
m
a
p
s
d
ep
ict
h
o
w
th
e
ea
r
th
'
s
s
u
r
f
ac
e
h
as
ch
a
n
g
ed
th
r
o
u
g
h
o
u
t
tim
e.
C
D
in
d
ata
is
co
n
s
id
er
ed
o
n
e
o
f
th
e
s
y
s
tem
a
tic
id
en
tific
atio
n
s
o
v
er
tim
e.
T
h
e
C
D
p
r
o
ce
s
s
h
as
p
r
o
f
o
u
n
d
s
ig
n
if
ican
ce
ac
r
o
s
s
d
iv
er
s
e
f
ield
s
,
in
clu
d
in
g
h
ea
lth
ca
r
e,
f
in
an
ce
,
en
v
ir
o
n
m
en
t
,
an
d
th
e
s
o
cial
s
cien
ce
s
.
B
y
id
en
tify
in
g
an
d
q
u
an
tify
in
g
ch
an
g
es,
C
D
h
elp
s
m
ak
e
v
alu
ab
le
d
ec
is
io
n
s
,
m
o
n
ito
r
tr
en
d
s
,
an
d
r
e
s
p
o
n
d
to
ev
o
l
v
in
g
cir
cu
m
s
tan
ce
s
ef
f
ec
tiv
ely
.
C
D
p
lay
s
a
v
ital
r
o
le
in
R
S
b
y
c
o
m
p
ar
in
g
th
e
s
p
ec
tr
al
a
n
d
s
p
a
tial
f
ea
tu
r
es
o
f
th
e
lan
d
,
a
n
d
v
eg
etatio
n
.
M
o
d
er
n
tech
n
iq
u
es
p
r
o
v
i
d
e
s
atellite
i
m
ag
es
with
h
ig
h
-
r
eso
lu
tio
n
im
ag
es,
an
d
m
a
n
y
C
D
tech
n
iq
u
es d
ep
e
n
d
o
n
th
eir
ac
cu
r
ac
y
[
6
]
–
[
8
]
.
C
D
an
d
class
if
icatio
n
o
f
s
atell
ite
im
ag
es
ar
e
p
o
wer
f
u
l
tech
n
i
q
u
es
th
at
p
r
o
v
id
e
v
alu
a
b
le
in
s
ig
h
ts
in
to
en
v
ir
o
n
m
en
tal
an
d
s
o
cieta
l
ch
an
g
es
o
v
er
tim
e.
W
ith
f
u
r
th
er
ad
v
an
ce
m
en
ts
in
im
ag
e
p
r
o
ce
s
s
in
g
an
d
ML
tech
n
iq
u
es,
th
e
m
eth
o
d
will
co
n
tin
u
e
to
b
e
a
n
ess
en
tial
to
o
l
f
o
r
r
esear
ch
er
s
an
d
p
r
ac
titi
o
n
e
r
s
in
v
ar
io
u
s
f
ield
s
.
Satellite
p
ictu
r
es
ar
e
cr
u
cial
in
m
u
ltip
le
ap
p
licatio
n
s
,
s
u
ch
as
d
is
aster
m
an
ag
em
en
t
an
d
en
v
i
r
o
n
m
e
n
tal
m
an
ag
em
en
t.
T
h
r
o
u
g
h
p
r
io
r
p
r
ed
ictio
n
,
R
S
ca
n
s
av
e
th
e
ea
r
th
f
r
o
m
n
atu
r
al
ca
lam
ities
an
d
wea
th
er
-
r
elate
d
th
r
ea
ts
,
an
d
t
h
er
e
is
a
n
ee
d
f
o
r
h
u
m
an
co
n
tr
ib
u
tio
n
s
in
id
e
n
tify
in
g
t
h
e
o
b
jects
in
th
e
im
ag
es,
an
d
th
e
s
am
e
s
tr
ateg
ies ar
e
u
s
ed
to
r
ep
r
esen
t
th
e
ea
r
th
p
r
ec
is
ely
[
9
]
.
Au
to
m
atic
C
D
tech
n
iq
u
es
b
etter
d
etec
t
m
u
ltip
le
ch
an
g
es
in
t
h
e
m
ap
s
th
an
in
th
e
2
D.
C
D
is
ess
en
tia
l
in
R
S
CD
in
r
ad
ar
r
ef
lectiv
ity
m
ea
s
u
r
em
en
ts
co
n
tam
in
ate
d
b
y
s
p
ec
k
le
n
o
is
e
.
Ad
v
an
ce
m
en
ts
in
th
e
g
lo
b
al
n
av
ig
atio
n
s
atellite
s
ch
em
e
ar
e
r
esp
o
n
s
ib
le
f
o
r
ex
p
a
n
d
in
g
th
e
u
tili
za
tio
n
o
f
R
S
to
p
r
o
cto
r
t
h
e
atm
o
s
p
h
e
r
e
a
n
d
th
e
ea
r
th
.
W
ith
ad
v
an
ce
d
s
y
s
tem
s
,
th
e
d
ata
ca
n
b
e
c
o
llected
ev
en
f
r
o
m
in
ac
ce
s
s
ib
le
an
d
r
em
o
te
ar
ea
s
b
y
o
f
f
er
in
g
g
lo
b
al
co
v
e
r
ag
e
[
1
0
]
.
T
h
e
im
p
r
o
v
em
e
n
t
in
th
e
R
S
t
ec
h
n
iq
u
e
h
as
led
to
im
p
r
o
v
ed
q
u
ality
an
d
clar
ity
o
f
th
e
R
S
im
ag
es
an
d
r
ed
u
ce
d
th
e
ef
f
o
r
t
o
f
o
b
tain
in
g
th
ese
i
m
ag
es,
an
d
h
u
m
a
n
s
ca
n
co
llec
t
lar
g
e
q
u
a
n
titi
es
o
f
im
ag
es
with
d
if
f
er
en
t
f
ea
t
u
r
e
s
an
d
r
eso
lu
tio
n
s
.
T
h
e
m
o
s
t
v
ital
in
f
o
r
m
atio
n
ca
n
b
e
co
llected
f
r
o
m
th
ese
im
ag
es,
an
d
th
ese
s
er
v
e
as
a
b
ase
f
o
r
u
n
d
er
s
tan
d
in
g
th
e
ea
r
th
'
s
s
y
s
tem
at
v
ar
io
u
s
lev
els
an
d
ca
n
b
e
u
s
ed
f
o
r
m
u
ltip
le
ap
p
licatio
n
s
lik
e
u
r
b
a
n
p
lan
n
i
n
g
,
ass
ess
in
g
clim
ate
ch
an
g
es,
an
d
m
o
n
ito
r
in
g
f
o
r
es
t r
eso
u
r
ce
s
[
1
1
]
.
T
h
e
an
aly
s
is
o
f
th
e
s
atellite
i
m
ag
es
ca
n
b
e
d
o
n
e
in
r
ea
l
-
ti
m
e
an
d
is
m
ad
e
p
o
s
s
ib
le
b
ec
au
s
e
o
f
th
e
av
ailab
ilit
y
o
f
m
an
y
s
atellites
ar
o
u
n
d
th
e
ea
r
t
h
.
Fo
r
R
S,
th
ese
class
if
ied
s
atellite
im
ag
es
p
r
o
v
id
e
m
a
n
y
b
en
ef
its
in
p
r
ed
ictio
n
s
[
1
2
]
.
T
h
e
im
ag
es
ca
p
tu
r
ed
b
y
u
n
m
an
n
ed
ae
r
ial
v
eh
icles
(
UAV)
s
u
f
f
er
f
r
o
m
p
r
o
b
lem
s
r
elate
d
to
th
e
b
ac
k
g
r
o
u
n
d
,
r
ed
u
ce
d
tar
g
ets,
a
n
d
h
i
d
d
en
tar
g
e
ts
,
wh
ich
r
esu
lts
in
r
ed
u
ce
d
ac
cu
r
ac
y
in
d
etec
tio
n
[
1
3
]
.
Ou
r
wo
r
k
p
r
o
p
o
s
es
a
m
eth
o
d
t
h
at
co
m
b
in
es
U
-
Net
a
n
d
y
o
u
o
n
ly
lo
o
k
o
n
ce
(
YOL
O)
s
tr
en
g
th
s
f
o
r
C
D
an
d
s
atellite
im
ag
er
y
class
if
ic
atio
n
.
W
e
m
eticu
lo
u
s
ly
ev
alu
ate
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
o
n
an
o
p
en
ly
a
v
ailab
le
d
ataset,
d
em
o
n
s
tr
atin
g
its
ef
f
ec
tiv
en
ess
in
d
etec
tin
g
a
n
d
class
if
y
in
g
ch
a
n
g
es
in
s
atellite
im
ag
er
y
.
T
h
is
th
o
r
o
u
g
h
e
v
alu
atio
n
in
s
till
s
co
n
f
id
en
ce
in
th
e
r
o
b
u
s
tn
ess
o
f
o
u
r
ap
p
r
o
ac
h
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
is
s
eg
m
en
t p
r
esen
ts
a
d
etail
ed
liter
atu
r
e
r
ev
iew
o
n
C
D
an
d
class
if
icatio
n
in
s
a
tellite
im
ag
er
y
u
s
in
g
d
ee
p
lear
n
in
g
(
DL
)
tech
n
iq
u
e
s
.
T
h
is
liter
atu
r
e
s
u
r
v
ey
aim
s
to
u
n
d
er
s
tan
d
th
e
p
r
esen
t
ad
v
an
ce
d
p
r
ac
tices
to
C
D
b
etter
.
Ver
y
h
ig
h
-
r
eso
lu
ti
o
n
im
ag
es a
r
e
u
s
ed
in
th
e
wo
r
s
t c
lim
atic
co
n
d
itio
n
s
with
o
u
t
b
ein
g
wo
r
r
ied
a
b
o
u
t
th
e
s
p
atial
d
etails.
T
h
e
tr
ad
itio
n
al
class
if
icatio
n
tech
n
iq
u
es n
ee
d
h
elp
to
h
a
n
d
le
th
e
co
m
p
le
x
ities
o
f
co
m
b
in
in
g
h
ig
h
-
r
eso
lu
tio
n
im
ag
es
with
a
h
eter
o
g
en
e
o
u
s
lan
d
s
ca
p
e.
T
h
e
s
o
lu
tio
n
u
s
es
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NN)
in
m
o
s
t
co
m
p
u
ter
v
is
io
n
ap
p
licatio
n
s
.
I
n
ad
d
itio
n
,
a
v
iab
le
p
latf
o
r
m
f
o
r
s
atellite
s
en
s
o
r
s
h
as
b
o
o
s
ted
th
is
g
r
o
wth
[
1
4
]
.
T
h
e
ap
p
licatio
n
s
o
f
R
S
ca
n
b
e
em
p
lo
y
ed
u
n
d
e
r
m
a
n
y
cir
cu
m
s
tan
ce
s
,
s
u
ch
as
ass
ess
in
g
d
am
a
g
e
af
ter
a
n
atu
r
al
d
is
aster
,
d
a
m
a
g
e
to
f
o
r
ests
af
ter
a
s
to
r
m
,
a
n
d
m
o
n
ito
r
i
n
g
g
lacie
r
m
eltin
g
an
d
d
ef
o
r
estatio
n
.
T
h
e
C
D
is
d
o
n
e
af
ter
co
m
p
ar
in
g
two
o
r
m
o
r
e
im
ag
es
tak
en
at
d
is
s
im
ilar
tim
es
an
d
d
ates
in
th
e
ex
ac
t
ter
r
estrial
lo
ca
tio
n
s
.
Var
io
u
s
m
eth
o
d
s
ex
is
t
to
co
llect
th
ese
im
ag
es,
b
u
t
s
atellite
im
ag
es
co
n
tin
u
o
u
s
ly
m
o
n
ito
r
th
e
en
tire
p
lan
et.
So
,
th
e
p
ict
u
r
es
o
f
th
e
s
atellites
ar
e
co
n
s
id
er
ed
a
v
alu
ab
le
s
o
u
r
ce
f
o
r
R
S
an
d
d
etec
tin
g
ch
an
g
es in
th
e
p
h
o
t
o
s
[
1
5
]
.
T
h
ese
d
ay
s
,
d
u
e
to
th
e
i
n
cr
e
ase
in
th
e
q
u
a
n
tity
o
f
ea
r
th
o
b
s
er
v
atio
n
s
atellites,
th
er
e
is
a
m
ass
iv
e
in
cr
ea
s
e
in
th
e
v
o
lu
m
e
o
f
th
e
d
ata
co
llected
,
an
d
th
is
i
n
cr
ea
s
es
th
e
lo
ad
b
ec
au
s
e
o
f
th
e
tr
an
s
m
is
s
io
n
b
an
d
wid
th
a
n
d
d
elay
s
in
c
o
m
m
u
n
icatio
n
.
A
n
au
to
m
atic
ch
an
g
e
id
e
n
tific
atio
n
s
y
s
tem
h
as
b
ee
n
d
ev
el
o
p
ed
u
s
in
g
th
e
ess
en
tial
f
ea
tu
r
es
o
f
DL
to
h
an
d
le
m
ass
iv
e
am
o
u
n
ts
o
f
d
ata.
T
h
e
ex
p
er
im
en
tal
r
e
s
u
lt
s
h
o
ws
th
at
th
e
s
y
s
tem
ac
h
iev
ed
an
ac
c
u
r
ac
y
o
f
9
1
.
9
5
%
[
1
6
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
C
h
a
n
g
e
d
etec
tio
n
a
n
d
cla
s
s
ifica
tio
n
o
f sa
tellite ima
g
es u
s
in
g
co
n
v
o
lu
tio
n
al
…
(
R
a
g
h
a
ve
n
d
r
a
S
r
in
iva
s
a
ia
h
)
331
Fo
r
m
o
s
t
R
S
p
u
r
p
o
s
es,
d
etec
t
io
n
o
f
ch
a
n
g
es
in
s
atellite
im
ag
es
is
th
e
m
o
s
t
im
p
o
r
tan
t
a
n
d
r
eq
u
ir
es
p
r
ec
is
e
b
o
u
n
d
a
r
y
d
etails.
Mo
s
t
ex
is
tin
g
m
eth
o
d
s
p
r
o
v
id
e
b
etter
f
ea
tu
r
e
ex
tr
ac
tio
n
t
h
r
o
u
g
h
p
ix
el
-
lev
el
co
m
p
ar
is
o
n
s
b
u
t
d
o
n
o
t
co
n
s
i
d
er
th
e
o
v
er
all
im
p
ac
t
th
at
b
lu
r
s
th
e
ed
g
es.
I
n
a
d
d
itio
n
to
t
h
is
,
th
ey
en
h
an
ce
t
h
e
co
m
p
lex
ity
o
f
th
e
c
o
m
p
u
tatio
n
.
T
o
r
e
p
o
r
t
t
h
ese
to
p
ics,
a
m
eth
o
d
ca
lled
f
ea
tu
r
e
e
n
h
an
c
em
en
t
an
d
f
ee
d
b
ac
k
n
etwo
r
k
(
FEFNet
)
is
p
r
o
jecte
d
f
o
r
C
D
to
im
p
r
o
v
e
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
an
d
p
r
o
v
id
es
v
al
u
ab
le
f
ee
d
b
ac
k
.
T
h
e
ex
p
er
im
en
tal
o
u
tco
m
e
ac
h
iev
e
d
an
ac
cu
r
ac
y
o
f
9
2
.
3
2
%
[
1
7
]
.
T
h
er
e
ar
e
v
ar
io
u
s
o
b
s
tacle
s
to
o
b
ject
r
ec
o
g
n
itio
n
in
s
atellite
im
ag
es,
in
clu
d
in
g
class
ch
an
g
es,
m
an
y
o
b
jects
in
m
o
tio
n
,
a
wid
e
r
a
n
g
e
o
f
item
s
izes,
lig
h
tin
g
,
an
d
a
b
u
s
y
b
ac
k
d
r
o
p
.
T
h
is
r
esear
ch
co
m
p
a
r
es
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
v
ar
io
u
s
DL
alg
o
r
ith
m
s
cu
r
r
en
tly
u
s
e
d
in
o
b
ject
d
etec
tio
n
in
s
atellite
im
ag
es.
U
s
in
g
f
r
am
ewo
r
k
s
b
ased
o
n
C
NN
⎯
lik
e
YOL
O,
f
aster
r
eg
io
n
-
b
ased
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
(
Fas
ter
R
C
NN)
,
s
atellite
im
ag
er
y
m
u
ltis
ca
le
r
a
p
id
d
etec
tio
n
with
win
d
o
wed
n
etwo
r
k
s
(
SIM
R
DW
N)
,
an
d
s
in
g
le
-
s
h
o
t
d
etec
to
r
(
SS
D)
⎯
an
d
a
co
llectio
n
o
f
s
atellite
p
h
o
to
s
is
co
n
s
tr
u
cted
to
co
n
d
u
ct
o
b
ject
r
ec
o
g
n
itio
n
.
Acc
o
r
d
in
g
to
th
e
d
ata,
SIM
R
DW
N
h
as
an
ac
cu
r
aten
ess
o
f
9
7
%
o
n
h
ig
h
-
q
u
ality
p
ict
u
r
es,
wh
e
r
ea
s
Fas
ter
R
C
NN
h
as
an
ac
cu
r
aten
ess
o
f
9
5
.
3
1
%
o
n
im
ag
es
with
a
s
tan
d
ar
d
r
eso
lu
tio
n
(
1
,
000
×
6
0
0
)
.
C
o
m
p
ar
e
d
to
SS
D,
YOL
Ov
3
h
as
an
ac
cu
r
aten
ess
o
f
9
4
.
2
0
%
at
s
tan
d
ar
d
r
eso
lu
tio
n
(
4
1
6
×
4
1
6
)
an
d
8
4
.
6
1
%
at
s
tan
d
ar
d
r
e
s
o
lu
tio
n
(
3
0
0
×
3
0
0
)
.
YOL
O
is
,
with
o
u
t
a
d
o
u
b
t,
t
h
e
b
est
in
s
p
ee
d
an
d
e
f
f
ec
tiv
en
ess
.
SIM
R
D
W
N
f
ail
s
in
r
ea
l
-
tim
e
s
u
r
v
eillan
ce
.
SIM
R
D
W
N
tak
es 5
to
1
0
3
m
il
lis
ec
o
n
d
s
to
co
m
p
lete
a
task
th
at
tak
es YO
L
O
1
7
0
to
1
9
0
m
il
lis
ec
o
n
d
s
[
1
8
]
.
I
n
R
S,
C
D
is
th
e
m
o
s
t
cr
itical
asp
ec
t.
A
C
D
m
ec
h
an
is
m
b
ased
o
n
t
h
e
u
n
s
u
p
er
v
is
ed
te
ch
n
iq
u
e
is
p
r
o
p
o
s
ed
b
y
o
p
tim
izin
g
th
e
p
r
o
d
u
ctio
n
an
d
a
n
aly
zin
g
t
h
e
d
if
f
er
en
t
im
ag
es.
T
h
e
weig
h
ted
v
ec
to
r
ca
lcu
latio
n
m
eth
o
d
is
u
s
ed
to
d
if
f
er
en
tia
te
b
etwe
en
th
e
v
ec
to
r
s
o
f
f
ea
tu
r
es
p
r
o
d
u
ce
d
b
y
th
e
clu
s
ter
in
g
.
At
last
,
th
e
Ma
r
k
o
v
tech
n
i
q
u
e
is
u
s
ed
to
g
en
er
ate
th
e
ch
an
g
e
m
ap
b
y
co
m
p
ar
in
g
it
with
th
e
n
eig
h
b
o
r
in
g
p
i
x
els.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
ac
h
iev
es a
n
ac
cu
r
ac
y
o
f
8
9
.
9
%
[
1
9
]
.
C
D
o
f
s
atellite
im
ag
es
i
s
an
u
n
av
o
id
a
b
le
s
tep
in
ea
r
th
o
b
s
er
v
atio
n
.
C
D
tech
n
iq
u
es
ar
e
h
elp
f
u
l
wh
e
n
ch
ar
ac
ter
izin
g
an
d
m
o
n
ito
r
in
g
u
r
b
an
g
r
o
wth
.
E
v
en
t
h
o
u
g
h
m
an
y
DL
m
ec
h
an
is
m
s
e
x
is
t
f
o
r
C
D,
m
a
n
y
ap
p
r
o
ac
h
es
f
ail
to
id
en
tify
th
e
ed
g
es
an
d
m
ai
n
tain
th
e
s
h
ap
e
o
f
t
h
e
ch
a
n
g
ed
ar
ea
s
.
A
DL
m
eth
o
d
ca
lle
d
u
r
b
an
ch
an
g
e
d
etec
tio
n
n
etwo
r
k
(
UC
DNe
t)
b
ased
o
n
th
e
en
c
o
d
er
-
d
ec
o
d
er
m
ec
h
an
is
m
is
d
ev
elo
p
e
d
to
ac
h
iev
e
b
etter
p
r
ed
ictio
n
with
o
u
t a
n
y
lo
s
s
in
th
e
im
ag
e
in
f
o
r
m
atio
n
.
UC
DNe
t a
ch
iev
es a
n
o
v
e
r
all
ac
cu
r
ac
y
o
f
8
9
.
2
1
%
[
2
0
]
.
Du
e
to
th
e
r
ap
id
tech
n
o
l
o
g
ica
l
r
ev
o
lu
tio
n
i
n
co
m
p
u
ter
v
is
io
n
,
h
ig
h
-
q
u
ality
s
atellite
im
ag
e
s
ar
e
v
ital
f
o
r
C
Ds.
E
m
p
lo
y
in
g
th
e
av
ai
lab
le
lim
ited
r
eso
u
r
ce
s
an
d
r
ed
u
cin
g
th
e
b
u
r
d
en
o
n
s
atellite
d
ev
ices
is
als
o
cr
u
cial.
A
s
ca
le
-
awa
r
e
p
r
u
n
in
g
f
r
am
ewo
r
k
(
SAPF
)
is
p
r
o
p
o
s
ed
to
r
ed
u
ce
th
e
co
m
p
lex
iti
es
an
d
m
an
ag
e
th
e
r
ep
r
esen
tatio
n
q
u
ality
.
T
o
in
itiate,
th
e
co
n
v
o
l
u
tio
n
al
lay
e
r
i
n
o
b
ject
r
ec
o
g
n
itio
n
is
alien
ated
in
to
two
g
r
o
u
p
s
:
s
in
g
le
-
s
ca
le
attr
ib
u
te
d
e
p
ictio
n
u
s
in
g
s
in
g
le
v
alu
e
b
r
ea
k
d
o
wn
an
d
m
u
ltis
ca
le
attr
ib
u
te
d
ep
ictio
n
u
s
ed
f
o
r
o
p
tim
izin
g
attr
ib
u
tes.
T
h
e
ex
p
er
im
en
tal
r
esu
lt
f
o
u
n
d
th
at
SAPF
r
ed
u
ce
d
th
e
p
ar
am
eter
s
an
d
f
lo
atin
g
p
o
in
t
o
p
er
atio
n
s
(
FLOPs
)
,
an
d
m
o
r
e
im
p
o
r
tan
tly
,
th
e
m
o
d
el'
s
ef
f
icien
cy
was g
r
ea
tly
im
p
r
o
v
e
d
[
2
1
]
.
An
y
C
D
m
et
h
o
d
co
m
b
i
n
i
n
g
th
e
m
e
th
o
d
o
f
f
e
at
u
r
e
e
x
t
r
a
ct
io
n
a
n
d
ML
ca
n
e
f
f
ec
ti
v
el
y
am
ass
th
e
in
f
o
r
m
at
io
n
c
o
m
p
ar
e
d
to
a
m
an
u
a
l
m
e
th
o
d
.
Us
in
g
m
a
n
u
al
m
et
h
o
d
s
ca
n
n
o
t
e
n
s
u
r
e
ac
c
u
r
a
cy
.
A
n
ew
wa
y
f
o
r
C
D
is
p
r
o
p
o
s
e
d
b
as
ed
o
n
t
h
e
f
u
s
io
n
o
f
m
u
lt
ip
le
f
e
at
u
r
es,
a
n
d
t
h
e
t
ec
h
n
i
q
u
e
is
ca
l
le
d
De
m
p
s
te
r
-
S
h
a
f
e
r
(
D
-
S
)
ev
i
d
en
ce
th
e
o
r
y
.
T
h
e
m
et
h
o
d
f
in
d
s
t
h
e
d
if
f
e
r
en
ce
s
in
t
h
e
im
a
g
es
b
ase
d
o
n
s
i
m
il
ar
i
ty
i
n
t
h
e
s
tr
u
ct
u
r
es
.
S
am
p
l
es
ar
e
s
ele
ct
ed
b
as
ed
o
n
s
p
ec
i
f
i
c
r
u
les
,
a
n
d
s
eg
m
e
n
t
ati
o
n
is
a
p
p
lie
d
to
ex
p
an
d
t
h
e
r
el
ia
b
il
it
y
o
f
th
e
s
a
m
p
les
.
T
h
e
r
es
u
lts
a
r
e
th
e
n
u
s
e
d
t
o
o
b
ta
in
th
e
r
es
u
lt
.
T
h
e
e
x
p
e
r
i
m
e
n
t
al
o
u
tc
o
m
es
f
o
u
n
d
th
at
th
e
p
r
o
j
ec
t
ed
wo
r
k
ac
h
ie
v
es a
n
ac
c
u
r
at
e
n
ess
o
f
9
0
.
7
6
%
u
s
in
g
t
h
e
a
v
e
r
a
g
e
s
t
r
u
c
tu
r
a
l si
m
i
la
r
it
y
i
n
d
e
x
m
e
asu
r
e
(
AV
E
-
SS
I
M
)
m
et
h
o
d
[
2
2
]
.
T
h
e
C
D
is
th
e
p
r
o
ce
s
s
o
f
i
d
e
n
tify
in
g
t
h
e
ch
an
g
es
in
th
e
g
r
o
u
n
d
im
ag
e
p
air
s
af
ter
co
m
p
ar
in
g
.
T
h
e
co
m
p
ar
is
o
n
d
o
n
e
at
t
h
e
s
ce
n
e,
o
b
ject,
an
d
p
ix
el
lev
els
is
v
ital
s
in
ce
it
p
r
o
v
id
es
th
e
s
em
an
tic
d
etails
an
d
r
eq
u
ir
es
m
o
n
ito
r
in
g
u
r
b
an
ar
e
a
ch
an
g
e.
T
h
e
ex
is
tin
g
au
t
o
m
atic
s
ce
n
e
-
lev
el
C
D
u
s
e
s
m
id
-
lev
el
an
d
lo
w
-
lev
el
attr
ib
u
tes
to
ex
tr
ac
t
ch
an
g
es
b
etwe
en
im
ag
es
an
d
f
ails
to
u
n
co
v
er
th
e
h
id
d
e
n
in
f
o
r
m
atio
n
.
A
n
o
v
el
au
to
m
atic
C
D
m
eth
o
d
at
t
h
e
b
i
n
ar
y
s
ce
n
e
lev
el
is
p
r
o
p
o
s
ed
to
h
an
d
le
th
e
m
e
n
tio
n
ed
p
r
o
b
lem
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
u
s
es
v
is
u
al
g
eo
m
et
r
y
g
r
o
u
p
(
VGG
)
-
1
6
at
th
e
f
ir
s
t
lev
el
f
o
r
p
r
e
-
tr
ain
in
g
,
d
ec
is
io
n
tr
ee
(
DT
)
f
o
r
p
ix
el
-
lev
el
class
if
icatio
n
in
th
e
s
ec
o
n
d
le
v
el,
tr
ain
in
g
s
am
p
les
at
th
e
s
ce
n
e
lev
el
ar
e
co
llected
at
th
e
th
ir
d
lev
el,
an
d
a
b
in
ar
y
s
ce
n
e
-
le
v
el
ch
an
g
e
m
ap
is
g
en
er
ated
at
th
e
last
.
T
h
e
ex
p
er
im
e
n
tal
r
esu
lts
f
o
u
n
d
th
at
th
e
p
r
o
p
o
s
ed
m
eth
o
d
ac
h
iev
es a
n
ac
cu
r
ac
y
o
f
9
2
.
1
7
%
[
2
3
]
.
T
h
e
C
D
is
th
e
m
o
s
t
cr
u
cial
te
ch
n
iq
u
e
to
a
n
aly
ze
th
e
ch
an
g
es
in
h
ig
h
-
d
ef
in
itio
n
im
ag
es.
I
d
en
tify
in
g
th
e
m
in
o
r
a
n
d
th
e
s
ig
n
if
ican
t
ch
an
g
es
an
d
p
r
o
d
u
cin
g
an
ac
c
u
r
ate
C
D
r
esu
lt
is
v
ital.
T
o
h
a
n
d
le
th
e
m
ajo
r
an
d
th
e
m
in
o
r
ch
a
n
g
es,
a
d
ee
p
s
u
p
er
v
is
ed
d
u
al
d
is
cr
im
in
ati
v
e
m
etr
ic
n
etwo
r
k
is
tr
ain
e
d
f
o
r
h
a
n
d
lin
g
th
e
bi
-
tem
p
o
r
al
im
ag
es.
T
h
e
p
r
o
p
o
s
ed
n
etwo
r
k
s
elec
ts
th
e
lo
w
-
s
tag
e
f
ea
tu
r
e
a
n
d
co
n
v
er
ts
it
i
n
to
g
lo
b
al
f
ea
t
u
r
es,
wh
ich
ar
e
m
o
r
e
r
eliab
le
an
d
h
ea
lth
ier
.
T
h
e
d
is
tan
ce
m
ea
s
u
r
e
is
u
s
ed
to
id
en
tify
th
e
d
if
f
er
en
ce
s
am
o
n
g
th
e
p
air
s
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
p
er
f
o
r
m
s
b
etter
c
o
m
p
a
r
ed
to
m
an
y
ex
is
tin
g
m
eth
o
d
s
[
2
4
]
.
T
h
e
c
h
a
r
a
c
t
e
r
is
t
i
cs
o
f
t
h
e
R
S
im
a
g
e
s
d
e
p
e
n
d
o
n
t
h
e
c
l
o
u
d
s
,
m
i
s
t
,
c
a
p
t
u
r
i
n
g
t
i
m
e
,
a
n
d
w
e
a
th
e
r
;
a
n
d
w
i
l
l
b
e
a
c
c
o
u
n
t
a
b
l
e
f
o
r
s
u
b
s
t
a
n
ti
a
l
s
p
e
c
t
r
a
l
m
o
d
i
f
i
ca
t
i
o
n
s
.
T
h
e
m
e
n
t
i
o
n
e
d
f
a
c
t
o
r
s
a
f
f
e
c
t
t
h
e
a
c
c
u
r
a
c
y
o
f
t
h
e
C
D
t
e
c
h
n
i
q
u
e
s
.
A
n
e
w
s
c
h
e
m
e
b
a
s
e
d
o
n
d
a
t
a
a
u
g
m
e
n
t
a
t
i
o
n
i
s
p
r
o
j
e
c
t
e
d
t
o
i
m
p
r
o
v
e
a
c
c
u
r
a
c
y
.
T
w
o
s
i
m
u
l
a
t
i
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
329
-
3
3
7
332
m
e
t
h
o
d
s
,
m
o
s
a
i
c
,
a
n
d
h
a
z
e
s
im
u
l
a
t
i
o
n
,
ai
m
t
o
c
r
e
a
te
n
e
w
s
a
m
p
l
e
s
t
o
c
o
n
d
u
c
t
d
a
t
a
a
u
g
m
e
n
t
a
t
i
o
n
.
T
h
e
l
at
e
s
t
s
a
m
p
l
es
a
r
e
m
i
x
e
d
w
it
h
t
h
e
o
r
ig
i
n
a
l
s
a
m
p
l
e
s
a
n
d
f
e
d
i
n
t
o
D
L
f
o
r
t
r
a
i
n
i
n
g
.
T
h
e
e
x
p
e
r
i
m
e
n
t
al
r
e
s
u
l
ts
s
h
o
w
e
d
t
h
a
t
t
h
e
p
r
o
p
o
s
e
d
m
o
d
e
l
c
o
u
l
d
h
a
n
d
l
e
t
h
e
m
is
t
r
e
g
i
o
n
b
e
tt
e
r
t
h
a
n
ex
i
s
t
i
n
g
m
et
h
o
d
s
,
w
it
h
a
n
a
c
c
u
r
a
c
y
o
f
8
4
.
6
1
%
[
2
5
]
.
C
D
in
v
o
lv
es
a
co
m
p
ar
is
o
n
o
f
im
ag
es
ca
p
tu
r
ed
at
d
is
s
im
ilar
tim
es.
C
u
r
r
en
t
C
D
tech
n
iq
u
es
f
o
cu
s
o
n
r
ec
o
r
d
e
d
p
ictu
r
es
an
d
d
o
n
o
t
c
o
n
s
id
er
th
e
task
s
p
o
s
ed
b
y
u
n
r
eg
is
ter
ed
p
air
s
o
f
im
ag
es.
L
ac
k
o
f
tr
ain
in
g
lead
s
to
n
o
is
y
o
u
tco
m
es.
T
o
p
u
t
an
en
d
to
th
e
m
en
tio
n
e
d
is
s
u
es,
a
n
ew
m
eth
o
d
b
ased
o
n
C
NN
an
d
g
en
er
ativ
e
ad
v
er
s
ar
ial
n
etwo
r
k
(
GAN
)
is
p
r
o
p
o
s
ed
,
wh
ich
au
to
m
ati
ca
lly
co
llects
d
ata
f
r
o
m
m
atch
in
g
r
eg
io
n
s
o
f
u
n
r
eg
is
ter
ed
im
ag
es
a
n
d
a
p
p
li
es
C
D
o
n
th
e
ex
tr
ac
te
d
r
e
g
io
n
,
an
d
C
NN
is
u
s
ed
to
co
llec
t
in
f
o
r
m
atio
n
f
r
o
m
u
n
r
eg
is
ter
ed
im
ag
es
f
o
llo
we
d
b
y
f
ea
tu
r
e
m
ap
p
in
g
to
id
e
n
tify
th
e
m
atch
in
g
r
e
g
io
n
s
.
Fro
m
th
e
ex
p
er
im
en
tal
an
aly
s
is
,
th
e
p
r
o
p
o
s
ed
m
eth
o
d
ac
h
iev
es 4
.
4
6
% h
ig
h
er
ac
cu
r
a
cy
th
an
th
e
ex
is
tin
g
m
eth
o
d
s
[
2
6
]
.
T
h
is
p
ap
er
p
r
o
p
o
s
es
a
n
o
v
el
C
NN
m
o
d
el
th
at
in
g
ests
m
an
y
m
u
lti
-
ch
an
n
el
3
D
d
ata
f
r
o
m
s
atellite
im
ag
er
y
an
d
lo
ca
l
3
D
m
a
p
p
in
g
d
ata.
T
h
e
p
r
o
p
o
s
ed
C
NN
m
o
d
el
h
as
4
3
la
y
er
s
.
L
ea
r
n
in
g
h
as
b
ee
n
ca
teg
o
r
ized
in
to
f
o
u
r
c
o
m
m
o
n
ty
p
es
o
f
r
o
o
f
in
g
m
ater
ials
,
an
d
b
ec
au
s
e
it
f
r
eq
u
en
tly
d
is
p
lay
s
p
atter
n
s
s
im
ilar
to
th
o
s
e
s
u
r
r
o
u
n
d
in
g
it,
th
e
p
ap
er
ex
h
ib
its
th
e
m
o
d
el
t
o
class
if
y
f
e
atu
r
es
ac
cu
r
ately
.
C
o
m
p
ar
ed
to
Go
o
g
leNe
t,
t
h
e
s
u
g
g
ested
m
o
d
el’
s
lear
n
in
g
o
u
tco
m
es r
ev
ea
led
a
9
% im
p
r
o
v
em
en
t in
m
ater
ial
ca
teg
o
r
izatio
n
ac
cu
r
ac
y
[
2
7
]
.
Mo
n
ito
r
in
g
u
r
b
an
izatio
n
an
d
ag
r
icu
ltu
r
al
lan
d
an
d
u
p
d
ati
n
g
g
eo
s
p
atial
d
atab
ases
th
r
o
u
g
h
C
D
in
s
atellite
im
ag
es
is
v
er
y
im
p
o
r
tan
t.
DL
-
b
ased
C
D
tech
n
iq
u
es
m
ain
ly
f
o
cu
s
o
n
tex
tu
r
e
an
d
c
o
lo
r
an
d
f
ac
e
m
a
n
y
ch
allen
g
es
d
u
e
t
o
th
e
b
ac
k
g
r
o
u
n
d
r
esem
b
la
n
ce
in
th
e
s
u
r
r
o
u
n
d
in
g
r
eg
io
n
s
.
I
n
a
d
d
itio
n
,
r
ed
u
cin
g
th
e
d
o
wn
s
am
p
lin
g
o
f
t
h
e
im
ag
es
m
ay
lead
t
o
a
lo
s
s
o
f
s
p
atial
d
ata.
A
n
o
v
el
n
etwo
r
k
,
atte
n
tio
n
-
b
ased
f
ea
tu
r
e
d
if
f
er
en
tial
en
h
an
ce
m
en
t
n
etwo
r
k
(
AFDE
-
NE
T
)
,
is
p
r
o
j
ec
ted
to
m
in
im
ize
d
ata
lo
s
s
an
d
u
s
es
a
d
ee
p
s
u
p
er
v
is
io
n
m
o
d
el
co
m
b
in
e
d
with
an
en
s
em
b
le
s
p
atial
ch
an
n
el
to
h
an
d
le
th
ese.
T
h
e
p
r
o
je
cted
m
o
d
el
ac
h
iev
es
an
ac
cu
r
ac
y
o
f
9
4
.
3
% wh
en
ap
p
lied
to
th
e
E
g
y
p
t
b
u
ild
in
g
ch
an
g
e
d
etec
tio
n
(
E
GY
-
BCD
)
d
ataset
[
2
8
]
.
C
u
r
r
en
t
ea
r
th
o
b
s
er
v
atio
n
d
at
a
p
r
o
v
id
e
q
u
alitativ
e
an
d
q
u
a
n
titativ
e
in
f
o
r
m
atio
n
co
m
p
ar
e
d
to
ea
r
lier
lan
d
-
r
elate
d
s
u
r
v
ey
s
.
R
S
o
f
f
er
s
d
ata
r
elate
d
to
p
o
liti
ca
l,
ec
o
n
o
m
ic,
an
d
s
cien
tific
d
ata.
M
an
y
ch
allen
g
es
ar
e
en
co
u
n
ter
e
d
wh
e
n
class
if
y
in
g
s
atellite
im
ag
es
an
d
h
a
n
d
lin
g
th
ese
im
ag
es;
s
ix
ML
tech
n
iq
u
es
s
u
ch
as
DT
,
r
an
d
o
m
f
o
r
est
(
R
F),
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
e
(
SVM)
,
class
if
icatio
n
an
d
r
eg
r
ess
io
n
tr
ee
s
(
C
AR
T
)
,
m
in
im
u
m
d
is
tan
ce
(
MD
)
,
a
n
d
g
r
ad
ien
t
t
r
ee
b
o
o
s
t
(
GT
B
)
.
B
ased
o
n
th
e
ex
p
er
im
e
n
t
r
esu
lt
it
was
f
o
u
n
d
th
at
an
ac
c
u
r
ac
y
o
f
9
3
%
was
ac
h
iev
ed
u
s
in
g
MD
[
2
9
]
.
T
h
is
liter
atu
r
e
s
u
r
v
ey
p
r
o
v
id
es
a
s
o
lid
f
o
u
n
d
atio
n
f
o
r
th
e
p
r
o
p
o
s
ed
wo
r
k
an
d
s
er
v
es
as
a
g
u
id
e
f
o
r
id
en
tify
in
g
th
e
m
o
s
t
p
r
o
m
is
in
g
ap
p
r
o
ac
h
es
an
d
tech
n
iq
u
es
f
o
r
d
ev
elo
p
i
n
g
a
s
y
s
tem
f
o
r
CD
an
d
class
if
icatio
n
u
s
in
g
U
-
Net
a
n
d
YOL
O
o
b
ject
d
etec
tio
n
.
3.
M
E
T
H
O
D
Her
e,
we
p
r
esen
t
t
h
e
wo
r
k
o
f
d
ev
elo
p
in
g
a
s
y
s
tem
f
o
r
C
D
a
n
d
class
if
icatio
n
u
s
in
g
U
-
Net
an
d
YOL
O
o
b
ject
d
etec
tio
n
.
T
h
e
m
ain
o
b
jectiv
e
o
f
th
is
r
esear
ch
is
to
p
r
o
p
o
s
e
an
ap
p
r
o
ac
h
th
at
co
m
b
i
n
es
th
e
s
tr
en
g
th
s
o
f
U
-
Net
an
d
YOL
O
to
d
etec
t
a
n
d
ac
cu
r
ately
class
if
y
ch
an
g
e
s
in
s
atel
lite
im
ag
er
y
.
B
y
co
m
b
in
in
g
th
e
s
tr
en
g
th
s
o
f
th
ese
two
ML
m
o
d
els,
o
u
r
ap
p
r
o
ac
h
ca
n
ef
f
ec
tiv
el
y
id
e
n
tify
an
d
class
if
y
ch
an
g
es
b
et
wee
n
two
o
r
m
o
r
e
im
ag
es o
f
th
e
s
am
e
ar
ea
tak
en
at
d
if
f
er
en
t tim
es.
T
h
e
f
r
a
m
ew
o
r
k
f
o
r
th
e
p
r
o
p
o
s
ed
w
o
r
k
is
s
h
o
w
n
i
n
Fi
g
u
r
e
1
.
Fi
r
s
t
,
i
m
a
g
e
p
a
ir
s
o
f
s
at
elli
te
o
r
ae
r
ia
l
im
a
g
es
o
f
t
h
e
e
x
a
ct
lo
ca
ti
o
n
t
a
k
e
n
at
tw
o
d
i
f
f
e
r
e
n
t
ti
m
e
p
o
i
n
t
s
a
r
e
p
ass
e
d
to
a
p
r
e
p
r
o
ce
s
s
in
g
n
o
d
e
t
h
at
p
r
ep
ar
es
th
e
m
f
o
r
i
n
p
u
t
in
to
t
h
e
f
ir
s
t
C
NN
m
o
d
e
l
b
y
p
e
r
f
o
r
m
i
n
g
b
asi
c
i
m
a
g
e
p
r
o
c
ess
i
n
g
o
p
e
r
at
io
n
s
a
n
d
f
o
r
m
at
ti
n
g
t
o
co
m
p
ati
b
l
e
i
n
p
u
t
f
o
r
m
at
.
On
c
e
f
e
d
i
n
t
o
t
h
e
1
st
C
NN
,
t
h
e
i
m
a
g
es
ar
e
c
o
m
p
ar
ed
,
a
n
d
t
h
e
d
et
ec
te
d
c
h
a
n
g
e
is
h
i
g
h
li
g
h
te
d
i
n
t
h
e
g
e
n
er
ate
d
f
e
atu
r
e
m
a
p
,
w
h
i
ch
h
i
g
h
li
g
h
ts
th
e
a
ct
u
a
l
ch
an
g
e
in
t
h
e
r
e
g
i
o
n
wit
h
i
n
t
h
e
i
m
a
g
e.
T
h
e
f
lo
w
o
f
d
ata
ac
r
o
s
s
th
e
p
r
o
p
o
s
ed
s
y
s
tem
is
s
h
o
wn
in
Fig
u
r
e
2
an
d
is
m
an
ag
e
d
th
r
o
u
g
h
w
e
l
l
-
d
e
f
i
n
e
d
p
i
p
e
l
i
n
es
w
h
e
r
e
th
e
s
u
p
p
o
r
t
i
n
g
f
u
n
c
t
i
o
n
s
e
n
s
u
r
e
t
h
e
c
o
r
r
e
c
t
d
at
a
r
e
a
c
h
es
t
h
e
r
i
g
h
t
p
l
a
c
es
a
t
t
h
e
r
i
g
h
t
t
i
m
e
.
Da
t
a
i
n
t
h
e
p
r
o
p
o
s
e
d
m
o
d
e
l
is
m
a
i
n
l
y
t
r
a
n
s
m
i
t
t
e
d
i
n
i
m
a
g
e
f
i
l
es
,
N
u
m
P
y
a
r
r
a
y
s
,
a
n
d
Py
T
o
r
c
h
o
b
j
e
c
t
s
s
u
ch
a
s
T
e
n
s
o
r
s
o
r
.
p
t
f
i
l
e
s
w
h
e
n
t
e
m
p
o
r
a
r
i
l
y
s
t
o
r
i
n
g
f
e
a
t
u
r
e
m
a
p
s
.
A
l
l
o
u
t
p
u
ts
a
r
e
s
t
o
r
e
d
i
n
w
e
l
l
-
es
t
a
b
li
s
h
ed
d
i
r
e
c
t
o
r
i
es
t
o
p
r
e
v
e
n
t
t
h
e
l
o
s
s
o
f
r
e
s
u
l
ts
.
T
h
e
s
tep
s
in
v
o
lv
ed
in
p
r
ep
r
o
ce
s
s
in
g
ar
e
as f
o
llo
ws:
i)
R
ea
d
th
e
in
p
u
t im
ag
es f
r
o
m
t
h
e
d
is
k
an
d
s
to
r
es th
em
in
m
em
o
r
y
.
ii)
I
n
p
u
t im
a
g
es a
r
e
d
iv
id
e
d
in
to
s
m
aller
p
atch
es o
r
tiles
o
f
s
ize
2
5
6
×
2
5
6
.
iii)
T
h
e
p
ix
el
v
alu
es
o
f
t
h
e
in
p
u
t
i
m
ag
es
ar
e
n
o
r
m
alize
d
to
th
e
r
an
g
e
[
0
,
1
]
u
s
in
g
m
in
-
m
a
x
n
o
r
m
aliza
tio
n
to
h
av
e
a
m
ea
n
o
f
0
.
5
a
n
d
a
s
tan
d
ar
d
d
e
v
iatio
n
o
f
0
.
5
.
iv
)
T
h
e
p
ix
el
-
wis
e
an
n
o
tatio
n
s
ar
e
in
teg
er
e
n
co
d
e
d
to
c
o
n
v
e
r
t
th
em
in
to
te
n
s
o
r
s
with
s
ize
[
n
B
atch
,
n
C
lass
,
h
eig
h
t,
wid
th
]
.
Pix
el
v
alu
es r
a
n
g
in
g
[
0
to
n
C
lass
-
1
]
.
v)
Data
au
g
m
en
tatio
n
tech
n
iq
u
es
s
u
ch
as
r
o
tatio
n
,
f
lip
p
in
g
,
a
n
d
zo
o
m
in
g
ar
e
ap
p
lied
to
in
cr
ea
s
e
th
e
s
ize
o
f
th
e
d
ataset
an
d
to
im
p
r
o
v
e
th
e
m
o
d
el’
s
r
o
b
u
s
tn
ess
.
v
i)
T
h
e
p
atch
es
ar
e
g
r
o
u
p
e
d
in
to
b
atch
es
o
f
a
s
p
ec
if
ied
s
ize
an
d
f
ed
in
t
o
th
e
U
-
Net
m
o
d
el
f
o
r
tr
ain
in
g
,
v
alid
atio
n
,
o
r
test
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
C
h
a
n
g
e
d
etec
tio
n
a
n
d
cla
s
s
ifica
tio
n
o
f sa
tellite ima
g
es u
s
in
g
co
n
v
o
lu
tio
n
al
…
(
R
a
g
h
a
ve
n
d
r
a
S
r
in
iva
s
a
ia
h
)
333
T
h
e
U
-
Net
m
o
d
el
is
im
p
lem
e
n
ted
u
s
in
g
th
e
Py
T
o
r
c
h
n
e
u
r
a
l
n
etwo
r
k
s
u
b
-
lib
r
a
r
y
an
d
h
as
an
in
p
u
t
s
ize
o
f
2
5
6
×
256
×
3
an
d
an
o
u
tp
u
t
s
ize
o
f
2
5
6
×
256
×
3
.
I
t
s
tar
ts
at
it
s
f
ir
s
t
lay
er
with
6
4
co
n
v
o
lu
ti
o
n
al
f
ilter
s
an
d
h
as
1
0
2
4
f
ilter
s
in
its
d
ee
p
est
5
th
lay
er
.
T
h
e
m
o
d
el
co
m
p
r
is
es
o
n
ly
co
n
v
o
lu
tio
n
al,
m
ax
p
o
o
lin
g
,
up
-
co
n
v
o
lu
tio
n
al,
an
d
d
r
o
p
o
u
t
lay
er
s
to
p
r
ev
en
t
o
v
er
f
itti
n
g
lay
er
s
with
th
e
L
ea
k
y
R
eL
U
ac
tiv
atio
n
f
u
n
ctio
n
,
wh
ich
h
as
a
s
lo
p
e
o
f
0
.
2
.
T
h
e
Ad
am
o
p
tim
izer
is
u
s
ed
to
o
p
tim
ize
th
e
m
o
d
el
with
a
lear
n
in
g
r
ate
o
f
0
.
0
0
2
,
an
d
th
e
lo
s
s
f
u
n
ctio
n
u
s
ed
is
t
h
e
cr
o
s
s
-
en
tr
o
p
y
lo
s
s
.
T
h
is
wo
r
k
'
s
C
D
a
s
p
ec
t
is
im
p
lem
en
ted
as
a
m
o
d
if
ied
in
f
er
en
cin
g
f
u
n
ctio
n
f
o
r
th
e
p
r
ev
i
o
u
s
ly
tr
ain
ed
s
em
an
tic
s
eg
m
en
tatio
n
m
o
d
el
.
T
h
e
f
u
n
ctio
n
f
ir
s
t
in
g
ests
th
e
b
ef
o
r
e
-
in
s
tan
ce
o
f
th
e
b
ite
m
p
o
r
al
p
air
o
f
in
p
u
t
im
ag
es
f
o
r
in
f
er
e
n
ce
an
d
p
er
f
o
r
m
s
th
e
co
n
tr
ac
tiv
e
p
h
ase
o
f
th
e
n
etwo
r
k
.
T
h
e
g
en
e
r
ated
f
ea
tu
r
e
m
ap
s
at
th
e
en
d
o
f
ea
ch
o
f
t
h
e
f
iv
e
lay
er
s
ar
e
th
en
s
av
ed
tem
p
o
r
ar
ily
a
s
Py
T
o
r
ch
o
b
jects.
T
h
e
f
u
n
cti
o
n
th
en
in
g
ests
th
e
af
ter
-
in
s
tan
ce
o
f
th
e
b
i
-
tem
p
o
r
al
p
air
o
f
in
p
u
t
im
ag
es
a
n
d
r
u
n
s
th
r
o
u
g
h
th
e
co
n
tr
ac
tiv
e
p
h
ase.
T
h
e
n
ewly
g
en
er
ated
f
ea
tu
r
e
m
a
p
s
ar
e
t
h
en
d
if
f
er
en
ce
d
f
r
o
m
th
e
p
r
e
v
io
u
s
ly
s
av
ed
f
ea
t
u
r
e
m
a
p
s
with
an
em
p
ir
ically
ca
lcu
lated
th
r
esh
o
ld
v
alu
e
f
o
r
ea
ch
lay
er
to
p
r
o
v
id
e
a
s
in
g
le
d
if
f
er
en
ce
f
e
atu
r
e
m
ap
f
o
r
th
e
ex
p
an
s
iv
e
p
h
ase
o
f
th
e
n
etwo
r
k
,
af
ter
wh
ich
th
e
r
e
s
u
ltan
t im
ag
e
is
a
p
r
ed
icted
s
e
m
an
tically
s
eg
m
en
ted
c
h
an
g
e
m
ap
.
Fig
u
r
e
1
.
Fra
m
ewo
r
k
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
Fig
u
r
e
2
.
Data
f
l
o
w
d
iag
r
am
I
mag
e
d
a
t
a
b
a
se
Pre
-
p
r
o
c
e
ssi
n
g
n
o
d
e
C
N
N
1
(
c
h
a
n
g
e
d
e
t
e
c
t
o
r
)
U
-
N
e
t
O
u
t
p
u
t
i
ma
g
e
s
I
n
t
e
r
me
d
i
a
r
y
p
r
o
c
e
ss
i
n
g
n
o
d
e
C
N
N
2
(
o
b
j
e
c
t
c
l
a
ssi
f
i
e
r
)
Y
O
LO
V
3
O
u
t
p
u
t
p
r
o
c
e
ss
i
n
g
n
o
d
e
D
a
t
a
b
a
s
e
o
f
bi
-
t
e
m
p
o
r
a
l
i
ma
g
e
s
Pre
-
p
r
o
c
e
ssi
n
g
o
p
e
r
a
t
i
o
n
s
N
e
u
r
a
l
n
e
t
w
o
r
k
I
n
t
e
r
med
i
a
r
y
p
r
o
c
e
ss
i
n
g
N
e
u
r
a
l
n
e
t
w
o
r
k
Pre
-
p
r
o
c
e
ssi
n
g
o
p
e
r
a
t
i
o
n
s
O
u
t
p
u
t
i
ma
g
e
s
a
n
d
t
e
m
p
o
r
a
r
y
s
t
o
r
e
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
329
-
3
3
7
334
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
s
em
an
tic
s
eg
m
en
tatio
n
m
o
d
el
u
s
ed
in
th
is
wo
r
k
was
tr
ain
ed
u
s
in
g
a
6
-
lab
el
a
n
d
a
3
-
lab
el
v
er
s
io
n
o
f
th
e
s
em
an
tic
s
eg
m
en
tatio
n
o
f
ae
r
ial
im
ag
e
r
y
d
at
aset
.
T
h
e
s
em
an
tic
s
eg
m
en
tatio
n
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
is
s
h
o
wn
in
Fig
u
r
e
3
.
W
h
en
tr
ain
ed
o
n
t
h
e
3
-
lab
el
v
er
s
io
n
o
f
th
e
d
ataset,
th
e
s
am
e
m
o
d
el
g
en
e
r
ated
a
s
lig
h
tly
lo
wer
p
er
f
o
r
m
a
n
ce
m
etr
ic,
lik
ely
d
u
e
to
class
im
b
a
lan
ce
.
T
h
e
m
o
d
el
t
r
ain
ed
with
th
e
6
-
la
b
el
d
ataset
ac
h
iev
ed
an
ac
cu
r
ac
y
s
co
r
e
o
f
o
v
er
8
5
%
a
n
d
a
m
ea
n
in
ter
s
ec
tio
n
o
v
er
u
n
io
n
o
f
0
.
6
2
0
.
T
h
e
m
o
d
el
tr
ain
in
g
m
etr
ics f
o
r
th
e
6
-
lab
el
v
e
r
s
io
n
,
as seen
in
Fig
u
r
e
4
,
in
d
icate
s
lig
h
tly
o
v
er
f
itti
n
g
at
p
lay
.
Figur
e 3. Se
m
a
nt
i
c s
egm
en
t
at
i
on r
es
u
l
t
Figur
e 4. Se
m
a
nt
i
c s
egm
en
t
at
i
on t
r
a
i
ni
n
g m
et
r
i
cs
T
h
e
C
D
o
p
er
atio
n
p
er
f
o
r
m
ed
b
y
th
e
p
r
ev
io
u
s
ly
d
is
cu
s
s
ed
Sem
an
tic
s
eg
m
en
tatio
n
m
o
d
el
g
en
er
ated
a
ch
an
g
e
m
a
p
d
u
e
t
o
th
e
d
if
f
er
e
n
ce
in
f
ea
tu
r
e
m
ap
s
o
f
th
e
tw
o
in
p
u
t
im
ag
es,
as
s
h
o
wn
i
n
Fig
u
r
e
5
.
D
u
e
to
th
e
ab
s
en
ce
o
f
m
u
lticlas
s
b
i
-
tem
p
o
r
al
p
h
o
t
o
s
in
th
e
tr
ain
in
g
d
a
taset,
m
an
u
ally
au
g
m
en
te
d
ch
an
g
e
is
in
tr
o
d
u
ce
d
in
to
th
e
test
im
ag
es,
an
d
th
e
P
C
C
2
s
co
r
e
o
f
ar
o
u
n
d
0
.
9
0
q
u
a
n
tifie
s
th
e
m
o
d
el’
s
p
er
f
o
r
m
a
n
ce
.
Fig
u
r
e
6
d
ep
icts
th
e
r
esu
lts
o
f
th
e
o
b
ject
d
etec
tio
n
m
o
d
el
u
s
in
g
YOL
Ov
3
b
y
h
ig
h
lig
h
tin
g
th
e
s
q
u
ar
e
b
o
x
wh
er
e
it
d
etec
ts
th
e
lab
eled
o
b
ject.
Figur
e 5. Cha
ng
e
m
ap
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
C
h
a
n
g
e
d
etec
tio
n
a
n
d
cla
s
s
ifica
tio
n
o
f sa
tellite ima
g
es u
s
in
g
co
n
v
o
lu
tio
n
al
…
(
R
a
g
h
a
ve
n
d
r
a
S
r
in
iva
s
a
ia
h
)
335
Fig
u
r
e
6
.
Ob
ject
d
etec
tio
n
r
es
u
lt
5.
CO
NCLU
SI
O
N
B
a
s
e
d
o
n
th
e
r
e
s
u
l
t
s
o
f
th
i
s
s
t
u
d
y
,
i
t
i
s
i
n
e
v
i
t
ab
l
e
t
h
a
t
m
u
l
t
i
c
l
a
s
s
s
em
a
n
t
i
c
s
e
g
m
en
ta
t
i
o
n
i
s
a
c
o
m
p
l
i
c
a
t
ed
co
m
p
u
t
er
v
i
s
i
o
n
p
r
o
b
l
e
m
.
T
h
i
s
w
o
r
k
a
ch
i
e
v
e
s
C
D
a
s
a
r
e
s
u
l
t
o
f
c
ar
e
f
u
l
f
e
a
tu
r
e
e
x
tr
a
c
t
io
n
an
d
f
e
a
t
u
r
e
d
i
f
f
e
r
en
c
e
c
a
l
c
u
la
t
i
o
n
f
r
o
m
th
e
s
e
m
a
n
t
i
c
s
e
g
m
en
t
at
i
o
n
p
r
o
c
e
s
s
w
i
t
h
a
d
e
c
en
t
l
ev
e
l
o
f
c
er
t
a
in
t
y
u
p
t
o
a
c
e
r
t
a
in
p
e
r
c
e
n
t
ag
e
o
f
p
ix
e
l
s
h
a
v
in
g
ch
a
n
g
e
.
T
h
e
m
o
r
e
s
i
g
n
if
i
c
a
n
t
t
h
e
ch
a
n
g
e
,
t
h
e
la
r
g
er
th
e
th
r
e
s
h
o
l
d
v
a
l
u
e
s
w
er
e
l
i
k
e
ly
t
o
in
d
u
c
e
f
a
l
s
e
p
o
s
i
t
iv
e
s
.
T
h
e
r
e
i
s
a
l
s
o
t
h
e
i
s
s
u
e
o
f
c
la
s
s
i
m
b
a
l
an
c
e
i
n
t
r
o
d
u
ce
d
to
t
h
e
r
e
d
u
c
t
i
o
n
i
n
l
ab
e
l
s
,
th
e
r
eb
y
cr
e
a
t
i
n
g
a
n
im
b
a
la
n
c
e
in
t
h
e
t
o
t
a
l
n
u
m
b
e
r
o
f
p
ix
e
l
s
o
f
a
p
ar
t
i
c
u
l
ar
c
l
a
s
s
.
T
h
e
l
a
b
e
l
r
e
d
u
c
t
i
o
n
w
a
s
p
er
f
o
r
m
e
d
i
n
th
e
b
e
s
t
i
n
t
er
e
s
t
o
f
t
h
e
m
o
d
e
l
’
s
p
e
r
f
o
r
m
a
n
c
e
c
o
n
ce
r
n
in
g
i
t
s
p
r
i
m
ar
y
o
b
j
e
c
t
iv
e
o
f
C
D
.
ACK
NO
WL
E
DG
M
E
N
T
S
W
e
wan
t to
th
an
k
C
HR
I
ST
Un
iv
er
s
ity
f
o
r
s
u
p
p
o
r
tin
g
u
s
in
co
m
p
letin
g
th
is
wo
r
k
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
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
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
R
a
g
h
a
v
e
n
d
r
a
S
r
i
n
i
v
a
s
a
i
a
h
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
San
to
s
h
K
u
m
ar
J
a
n
k
at
ti
✓
✓
✓
✓
✓
✓
✓
✓
Ma
n
ju
n
ath
R
am
an
n
a
L
am
an
i
✓
✓
✓
✓
✓
✓
✓
Nir
an
jan
a
Sh
r
av
an
ab
elag
o
la
J
in
ac
h
an
d
r
a
✓
✓
✓
✓
✓
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
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
329
-
3
3
7
336
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
D
a
t
a
a
v
a
i
l
a
b
i
l
i
ty
i
s
n
o
t
ap
p
li
c
a
b
l
e
t
o
th
i
s
p
a
p
er
a
s
n
o
n
e
w
d
a
t
a
w
e
r
e
cr
e
a
t
ed
o
r
a
n
a
l
y
z
ed
i
n
t
h
i
s
s
t
u
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
A
.
S
h
a
r
ma
e
t
a
l
.
,
“
A
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
t
e
c
h
n
i
q
u
e
s
f
o
r
l
a
n
d
s
l
i
d
e
s
p
r
e
d
i
c
t
i
o
n
u
s
i
n
g
sat
e
l
l
i
t
e
i
m
a
g
e
r
y
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
1
1
7
3
1
8
–
1
1
7
3
3
4
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
4
.
3
4
4
6
0
3
7
.
[
2
]
E.
K
a
r
a
k
ö
s
e
,
“
A
n
e
f
f
i
c
i
e
n
t
sa
t
e
l
l
i
t
e
i
mag
e
s
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
p
p
r
o
a
c
h
b
a
se
d
o
n
f
u
z
z
y
c
o
g
n
i
t
i
v
e
m
a
p
i
n
t
e
g
r
a
t
i
o
n
w
i
t
h
d
e
e
p
l
e
a
r
n
i
ng
mo
d
e
l
s
u
s
i
n
g
i
m
p
r
o
v
e
d
l
o
ss fu
n
c
t
i
o
n
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
1
4
1
3
6
1
–
1
4
1
3
7
9
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
4
.
3
4
6
1
8
7
1
.
[
3
]
Y
.
M
a
,
X
.
D
e
n
g
,
a
n
d
J.
W
e
i
,
“
La
n
d
u
se
c
l
a
ssi
f
i
c
a
t
i
o
n
o
f
h
i
g
h
-
r
e
so
l
u
t
i
o
n
m
u
l
t
i
s
p
e
c
t
r
a
l
s
a
t
e
l
l
i
t
e
i
m
a
g
e
s wi
t
h
f
i
n
e
-
g
r
a
i
n
e
d
mu
l
t
i
sc
a
l
e
n
e
t
w
o
r
k
s
a
n
d
s
u
p
e
r
p
i
x
e
l
p
o
s
t
p
r
o
c
e
ssi
n
g
,
”
I
EE
E
J
o
u
rn
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
Ap
p
l
i
e
d
E
a
rt
h
O
b
ser
v
a
t
i
o
n
s
a
n
d
Re
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
6
,
p
p
.
3
2
6
4
–
3
2
7
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
j
st
a
r
s.2
0
2
3
.
3
2
6
0
4
4
8
.
[
4
]
A
.
K
.
R
a
i
,
N
.
M
a
n
d
a
l
,
K
.
K
.
S
i
n
g
h
,
a
n
d
I
.
I
z
o
n
i
n
,
“
S
a
t
e
l
l
i
t
e
i
m
a
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
u
s
i
n
g
a
h
y
b
r
i
d
m
a
n
t
a
r
a
y
f
o
r
a
g
i
n
g
o
p
t
i
m
i
z
a
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
Bi
g
D
a
t
a
Mi
n
i
n
g
a
n
d
A
n
a
l
y
t
i
c
s
,
v
o
l
.
6
,
n
o
.
1
,
p
p
.
4
4
–
5
4
,
M
a
r
.
2
0
2
3
,
d
o
i
:
1
0
.
2
6
5
9
9
/
b
d
m
a
.
2
0
2
2
.
9
0
2
0
0
2
7
.
[
5
]
Y
.
Li
u
e
t
a
l
.
,
“
B
u
i
l
d
i
n
g
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
i
n
e
a
r
t
h
q
u
a
k
e
:
a
mu
l
t
i
sc
a
l
e
i
n
t
e
r
a
c
t
i
o
n
n
e
t
w
o
r
k
w
i
t
h
o
f
f
set
c
a
l
i
b
r
a
t
i
o
n
a
n
d
a
d
a
t
a
se
t
,
”
I
EEE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
G
e
o
sc
i
e
n
c
e
a
n
d
R
e
m
o
t
e
S
e
n
s
i
n
g
,
v
o
l
.
6
2
,
p
p
.
1
–
1
7
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
t
g
r
s.
2
0
2
4
.
3
4
3
8
2
9
0
.
[
6
]
B
.
S
h
a
h
,
A
.
G
u
p
t
a
,
a
n
d
S
.
P
a
u
l
,
“
A
P
S
O
-
S
V
M
-
b
a
s
e
d
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
a
l
g
o
r
i
t
h
m
f
o
r
r
e
m
o
t
e
se
n
s
i
n
g
o
p
t
i
c
a
l
i
ma
g
e
s,”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
5
4
2
2
9
–
5
4
2
3
7
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
4
.
3
3
8
7
9
4
0
.
[
7
]
L.
K
h
e
l
i
f
i
a
n
d
M
.
M
i
g
n
o
t
t
e
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
i
n
r
e
m
o
t
e
s
e
n
si
n
g
i
ma
g
e
s
:
c
o
m
p
r
e
h
e
n
si
v
e
r
e
v
i
e
w
a
n
d
m
e
t
a
-
a
n
a
l
y
si
s
,
”
I
E
EE
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
1
2
6
3
8
5
–
1
2
6
4
0
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
3
0
0
8
0
3
6
.
[
8
]
C
.
R
e
n
,
X
.
W
a
n
g
,
J.
G
a
o
,
X
.
Zh
o
u
,
a
n
d
H
.
C
h
e
n
,
“
U
n
s
u
p
e
r
v
i
se
d
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
i
n
sa
t
e
l
l
i
t
e
i
m
a
g
e
s
w
i
t
h
g
e
n
e
r
a
t
i
v
e
a
d
v
e
r
s
a
r
i
a
l
n
e
t
w
o
r
k
,
”
I
EEE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
G
e
o
sci
e
n
c
e
a
n
d
Re
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
5
9
,
n
o
.
1
2
,
p
p
.
1
0
0
4
7
–
1
0
0
6
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
t
g
r
s.2
0
2
0
.
3
0
4
3
7
6
6
.
[
9
]
A
.
H
a
mz
a
e
t
a
l
.
,
“
A
n
o
v
e
l
b
o
t
t
l
e
n
e
c
k
r
e
s
i
d
u
a
l
a
n
d
s
e
l
f
-
a
t
t
e
n
t
i
o
n
f
u
si
o
n
-
a
ssi
st
e
d
a
r
c
h
i
t
e
c
t
u
r
e
f
o
r
l
a
n
d
u
s
e
r
e
c
o
g
n
i
t
i
o
n
i
n
r
e
m
o
t
e
sen
s
i
n
g
i
ma
g
e
s
,
”
I
E
EE
J
o
u
r
n
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
A
p
p
l
i
e
d
E
a
rt
h
O
b
ser
v
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
7
,
p
p
.
2
9
9
5
–
3
0
0
9
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
j
s
t
a
r
s.
2
0
2
3
.
3
3
4
8
8
7
4
.
[
1
0
]
M
.
G
o
mr
o
k
i
,
M
.
H
a
s
a
n
l
o
u
,
a
n
d
J.
C
h
a
n
u
ss
o
t
,
“
A
u
t
o
m
a
t
i
c
3
D
m
u
l
t
i
p
l
e
b
u
i
l
d
i
n
g
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
m
o
d
e
l
b
a
se
d
o
n
e
n
c
o
d
e
r
–
d
e
c
o
d
e
r
n
e
t
w
o
r
k
u
si
n
g
h
i
g
h
l
y
u
n
b
a
l
a
n
c
e
d
r
e
m
o
t
e
s
e
n
s
i
n
g
d
a
t
a
s
e
t
s
,
”
I
EEE
J
o
u
r
n
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
A
p
p
l
i
e
d
E
a
rt
h
O
b
serv
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
s
i
n
g
,
v
o
l
.
1
6
,
p
p
.
1
0
3
1
1
–
1
0
3
2
5
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
j
st
a
r
s.2
0
2
3
.
3
3
2
8
5
6
1
.
[
1
1
]
Y
.
M
e
n
g
e
t
a
l
.
,
“
C
r
o
ss
-
d
o
m
a
i
n
l
a
n
d
c
o
v
e
r
c
l
a
ssi
f
i
c
a
t
i
o
n
o
f
r
e
m
o
t
e
se
n
si
n
g
i
ma
g
e
s
b
a
se
d
o
n
f
u
l
l
-
l
e
v
e
l
d
o
m
a
i
n
a
d
a
p
t
a
t
i
o
n
,
”
I
EE
E
J
o
u
rn
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
A
p
p
l
i
e
d
Ea
r
t
h
O
b
s
e
rv
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
7
,
p
p
.
1
1
4
3
4
–
1
1
4
5
0
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
j
st
a
r
s.
2
0
2
4
.
3
4
0
7
8
0
8
.
[
1
2
]
K
.
K
.
J
e
n
a
,
S
.
K
.
B
h
o
i
,
S
.
R
.
N
a
y
a
k
,
R
.
P
a
n
i
g
r
a
h
i
,
a
n
d
A
.
K
.
B
h
o
i
,
“
D
e
e
p
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
t
w
o
r
k
b
a
s
e
d
m
a
c
h
i
n
e
i
n
t
e
l
l
i
g
e
n
c
e
mo
d
e
l
f
o
r
sa
t
e
l
l
i
t
e
c
l
o
u
d
i
m
a
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
Bi
g
D
a
t
a
Mi
n
i
n
g
a
n
d
A
n
a
l
y
t
i
c
s
,
v
o
l
.
6
,
n
o
.
1
,
p
p
.
3
2
–
4
3
,
2
0
2
3
,
d
o
i
:
1
0
.
2
6
5
9
9
/
b
d
ma.
2
0
2
1
.
9
0
2
0
0
1
7
.
[
1
3
]
Y
.
L
i
,
D
.
F
e
n
g
,
Q
.
Z
h
a
n
g
,
a
n
d
S
.
L
i
,
“
H
R
D
-
Y
O
LO
X
b
a
se
d
i
n
su
l
a
t
o
r
i
d
e
n
t
i
f
i
c
a
t
i
o
n
a
n
d
d
e
f
e
c
t
d
e
t
e
c
t
i
o
n
m
e
t
h
o
d
f
o
r
t
r
a
n
sm
i
ssi
o
n
l
i
n
e
s,
”
I
EE
E
A
c
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
2
2
6
4
9
–
2
2
6
6
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ES
S
.
2
0
2
4
.
3
3
6
3
4
3
0
.
[
1
4
]
A
.
S
.
P
e
r
e
r
a
,
C
.
W
i
t
h
a
r
a
n
a
,
E
.
M
a
n
o
s,
a
n
d
A
.
K
.
Li
l
j
e
d
a
h
l
,
“
H
y
p
e
r
p
a
r
a
m
e
t
e
r
o
p
t
i
mi
z
a
t
i
o
n
f
o
r
l
a
r
g
e
-
sca
l
e
r
e
m
o
t
e
se
n
si
n
g
i
ma
g
e
a
n
a
l
y
si
s
t
a
s
k
s:
a
c
a
se
st
u
d
y
b
a
s
e
d
o
n
p
e
r
mafr
o
st
l
a
n
d
f
o
r
m
d
e
t
e
c
t
i
o
n
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
I
EE
E
Ac
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
4
3
0
6
2
–
4
3
0
7
7
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
4
.
3
3
7
9
1
4
2
.
[
1
5
]
J.
A
l
a
t
a
l
o
,
T
.
S
i
p
o
l
a
,
a
n
d
M
.
R
a
n
t
o
n
e
n
,
“
I
mp
r
o
v
e
d
d
i
f
f
e
r
e
n
c
e
i
m
a
g
e
s
f
o
r
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
c
l
a
ssi
f
i
e
r
s
i
n
S
A
R
i
m
a
g
e
r
y
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
G
e
o
sci
e
n
c
e
a
n
d
Re
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
6
1
,
p
p
.
1
–
1
4
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
t
g
r
s.
2
0
2
3
.
3
3
2
4
9
9
4
.
[
1
6
]
C
.
S
e
r
i
e
f
,
Y
.
G
h
e
l
a
ma
l
l
a
h
,
a
n
d
Y
.
B
e
n
t
o
u
t
o
u
,
“
D
e
e
p
-
l
e
a
r
n
i
n
g
-
b
a
se
d
s
y
st
e
m
f
o
r
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
o
n
b
o
a
r
d
e
a
r
t
h
o
b
serv
a
t
i
o
n
s
mal
l
sat
e
l
l
i
t
e
s,”
I
E
EE
J
o
u
rn
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
A
p
p
l
i
e
d
E
a
rt
h
O
b
s
e
rv
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
6
,
p
p
.
8
1
1
5
–
8
1
2
4
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
j
st
a
r
s.
2
0
2
3
.
3
2
8
4
9
1
9
.
[
1
7
]
Z.
J
i
a
n
g
e
t
a
l
.
,
“
F
e
a
t
u
r
e
e
n
h
a
n
c
e
m
e
n
t
a
n
d
f
e
e
d
b
a
c
k
n
e
t
w
o
r
k
f
o
r
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
i
n
r
e
mo
t
e
se
n
si
n
g
i
ma
g
e
s
,
”
I
EEE
G
e
o
s
c
i
e
n
c
e
a
n
d
Re
m
o
t
e
S
e
n
si
n
g
L
e
t
t
e
rs
,
v
o
l
.
2
2
,
p
p
.
1
–
5
,
2
0
2
5
,
d
o
i
:
1
0
.
1
1
0
9
/
l
g
r
s.2
0
2
5
.
3
5
3
6
1
6
4
.
[
1
8
]
A
.
Ta
h
i
r
e
t
a
l
.
,
“
A
u
t
o
ma
t
i
c
t
a
r
g
e
t
d
e
t
e
c
t
i
o
n
f
r
o
m
s
a
t
e
l
l
i
t
e
i
ma
g
e
r
y
u
si
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
S
e
n
s
o
rs
,
v
o
l
.
2
2
,
n
o
.
3
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
2
0
3
1
1
4
7
.
[
1
9
]
H
.
F
a
n
g
,
P
.
D
u
,
X
.
W
a
n
g
,
C
.
L
i
n
,
a
n
d
P
.
Ta
n
g
,
“
U
n
su
p
e
r
v
i
s
e
d
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
b
a
se
d
o
n
w
e
i
g
h
t
e
d
c
h
a
n
g
e
v
e
c
t
o
r
a
n
a
l
y
s
i
s
a
n
d
i
mp
r
o
v
e
d
M
a
r
k
o
v
r
a
n
d
o
m
f
i
e
l
d
f
o
r
h
i
g
h
sp
a
t
i
a
l
r
e
s
o
l
u
t
i
o
n
i
ma
g
e
r
y
,
”
I
E
EE
G
e
o
sc
i
e
n
c
e
a
n
d
R
e
m
o
t
e
S
e
n
s
i
n
g
L
e
t
t
e
rs
,
v
o
l
.
1
9
,
p
p
.
1
–
5
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
l
g
r
s.2
0
2
1
.
3
0
5
9
4
6
1
.
[
2
0
]
K
.
S
.
B
a
s
a
v
a
r
a
j
u
,
N
.
S
r
a
v
y
a
,
S
.
La
l
,
J.
N
a
l
i
n
i
,
C
.
S
.
R
e
d
d
y
,
a
n
d
F
.
D
e
l
l
’
A
c
q
u
a
,
“
U
C
D
N
e
t
:
a
d
e
e
p
l
e
a
r
n
i
n
g
mo
d
e
l
f
o
r
u
r
b
a
n
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
f
r
o
m
b
i
-
t
e
m
p
o
r
a
l
m
u
l
t
i
s
p
e
c
t
r
a
l
s
e
n
t
i
n
e
l
-
2
sa
t
e
l
l
i
t
e
i
ma
g
e
s
,
”
I
E
EE
T
r
a
n
sa
c
t
i
o
n
s
o
n
G
e
o
sci
e
n
c
e
a
n
d
Re
m
o
t
e
S
e
n
s
i
n
g
,
v
o
l
.
6
0
,
p
p
.
1
–
1
0
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
t
g
r
s.
2
0
2
2
.
3
1
6
1
3
3
7
.
[
2
1
]
Z.
H
u
e
t
a
l
.
,
“
S
c
a
l
e
-
a
w
a
r
e
p
r
u
n
i
n
g
f
r
a
mew
o
r
k
f
o
r
r
e
m
o
t
e
s
e
n
s
i
n
g
o
b
j
e
c
t
d
e
t
e
c
t
i
o
n
v
i
a
m
u
l
t
i
f
e
a
t
u
r
e
r
e
p
r
e
sen
t
a
t
i
o
n
,
”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
G
e
o
sc
i
e
n
c
e
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
6
3
,
p
p
.
1
–
1
5
,
2
0
2
5
,
d
o
i
:
1
0
.
1
1
0
9
/
t
g
r
s.2
0
2
5
.
3
5
4
5
4
1
6
.
[
2
2
]
J.
Z
h
a
o
,
S
.
Li
u
,
J.
W
a
n
,
M
.
Y
a
s
i
r
,
a
n
d
H
.
L
i
,
“
C
h
a
n
g
e
d
e
t
e
c
t
i
o
n
me
t
h
o
d
o
f
h
i
g
h
r
e
s
o
l
u
t
i
o
n
r
e
m
o
t
e
s
e
n
si
n
g
i
m
a
g
e
b
a
se
d
o
n
D
-
S
e
v
i
d
e
n
c
e
t
h
e
o
r
y
f
e
a
t
u
r
e
f
u
s
i
o
n
,
”
I
EE
E
Ac
c
e
ss
,
v
o
l
.
9
,
p
p
.
4
6
7
3
–
4
6
8
7
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
3
0
4
7
9
1
5
.
[
2
3
]
H
.
F
a
n
g
,
S
.
G
u
o
,
X
.
W
a
n
g
,
S
.
Li
u
,
C
.
L
i
n
,
a
n
d
P
.
D
u
,
“
A
u
t
o
ma
t
i
c
u
r
b
a
n
sce
n
e
-
l
e
v
e
l
b
i
n
a
r
y
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
b
a
s
e
d
o
n
a
n
o
v
e
l
samp
l
e
s
e
l
e
c
t
i
o
n
a
p
p
r
o
a
c
h
a
n
d
a
d
v
a
n
c
e
d
t
r
i
p
l
e
t
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
I
EE
E
T
ra
n
sa
c
t
i
o
n
s
o
n
G
e
o
sc
i
e
n
c
e
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
6
1
,
p
p
.
1
–
1
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
t
g
r
s.
2
0
2
3
.
3
2
3
5
9
1
7
.
[
2
4
]
X
.
Li
,
L.
Y
a
n
,
Y
.
Zh
a
n
g
,
a
n
d
N
.
M
o
,
“
S
D
M
N
e
t
:
a
d
e
e
p
-
s
u
p
e
r
v
i
se
d
d
u
a
l
d
i
s
c
r
i
mi
n
a
t
i
v
e
me
t
r
i
c
n
e
t
w
o
r
k
f
o
r
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
i
n
h
i
g
h
-
r
e
so
l
u
t
i
o
n
r
e
mo
t
e
se
n
si
n
g
i
mag
e
s,”
I
EE
E
G
e
o
s
c
i
e
n
c
e
a
n
d
Re
m
o
t
e
S
e
n
si
n
g
L
e
t
t
e
rs
,
v
o
l
.
1
9
,
p
p
.
1
–
5
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
l
g
r
s.2
0
2
2
.
3
2
1
6
6
2
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
C
h
a
n
g
e
d
etec
tio
n
a
n
d
cla
s
s
ifica
tio
n
o
f sa
tellite ima
g
es u
s
in
g
co
n
v
o
lu
tio
n
al
…
(
R
a
g
h
a
ve
n
d
r
a
S
r
in
iva
s
a
ia
h
)
337
[
2
5
]
Z.
W
a
n
g
,
D
.
L
i
u
,
Z.
W
a
n
g
,
X
.
Li
a
o
,
a
n
d
Q
.
Z
h
a
n
g
,
“
A
n
e
w
r
e
mo
t
e
s
e
n
si
n
g
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
d
a
t
a
a
u
g
me
n
t
a
t
i
o
n
met
h
o
d
b
a
s
e
d
o
n
mo
sa
i
c
si
mu
l
a
t
i
o
n
a
n
d
h
a
z
e
i
m
a
g
e
s
i
mu
l
a
t
i
o
n
,
”
I
EE
E
J
o
u
r
n
a
l
o
f
S
e
l
e
c
t
e
d
T
o
p
i
c
s
i
n
Ap
p
l
i
e
d
E
a
rt
h
O
b
s
e
rv
a
t
i
o
n
s
a
n
d
R
e
m
o
t
e
S
e
n
si
n
g
,
v
o
l
.
1
6
,
p
p
.
4
5
7
9
–
4
5
9
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
j
s
t
a
r
s.
2
0
2
3
.
3
2
6
9
7
8
4
.
[
2
6
]
J.
D
u
,
X
.
Z
h
o
u
,
a
n
d
H
.
C
h
e
n
,
“
En
h
a
n
c
e
d
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
i
n
u
n
r
e
g
i
st
e
r
e
d
i
m
a
g
e
s
w
i
t
h
C
N
N
s
a
n
d
a
t
t
e
n
t
i
o
n
G
A
N
s,”
I
EEE
G
e
o
sc
i
e
n
c
e
a
n
d
Re
m
o
t
e
S
e
n
si
n
g
L
e
t
t
e
rs
,
v
o
l
.
2
1
,
p
p
.
1
–
5
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
l
g
r
s.
2
0
2
4
.
3
4
3
4
7
3
0
.
[
2
7
]
J.
K
i
m,
H
.
B
a
e
,
H
.
K
a
n
g
,
a
n
d
S
.
G
.
Le
e
,
“
C
N
N
a
l
g
o
r
i
t
h
m
f
o
r
r
o
o
f
d
e
t
e
c
t
i
o
n
a
n
d
m
a
t
e
r
i
a
l
c
l
a
ss
i
f
i
c
a
t
i
o
n
i
n
s
a
t
e
l
l
i
t
e
i
m
a
g
e
s,”
El
e
c
t
r
o
n
i
c
s
,
v
o
l
.
1
0
,
n
o
.
1
3
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s
1
0
1
3
1
5
9
2
.
[
2
8
]
S
.
H
o
l
a
i
l
,
T.
S
a
l
e
h
,
X
.
X
i
a
o
,
a
n
d
D
.
L
i
,
“
A
F
D
E
-
N
e
t
:
b
u
i
l
d
i
n
g
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
u
s
i
n
g
a
t
t
e
n
t
i
o
n
-
b
a
s
e
d
f
e
a
t
u
r
e
d
i
f
f
e
r
e
n
t
i
a
l
e
n
h
a
n
c
e
me
n
t
f
o
r
sa
t
e
l
l
i
t
e
i
ma
g
e
r
y
,
”
I
EE
E
G
e
o
s
c
i
e
n
c
e
a
n
d
Re
m
o
t
e
S
e
n
si
n
g
L
e
t
t
e
rs
,
v
o
l
.
2
0
,
p
p
.
1
–
5
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
l
g
r
s.2
0
2
3
.
3
2
8
3
5
0
5
.
[
2
9
]
H
.
O
u
c
h
r
a
,
A
.
B
e
l
a
n
g
o
u
r
,
a
n
d
A
.
Er
r
a
i
ssi
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
f
o
r
sat
e
l
l
i
t
e
i
ma
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
u
s
i
n
g
G
o
o
g
l
e
E
a
r
t
h
e
n
g
i
n
e
a
n
d
l
a
n
d
s
a
t
s
a
t
e
l
l
i
t
e
d
a
t
a
:
M
o
r
o
c
c
o
c
a
se
st
u
d
y
,
”
I
E
E
E
Ac
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
7
1
1
2
7
–
7
1
1
4
2
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
2
9
3
8
2
8
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Ra
g
h
a
v
e
n
d
r
a
S
r
in
i
v
a
sa
i
a
h
is
c
u
rre
n
tl
y
wo
r
k
in
g
a
s
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
t
h
e
De
p
a
rtme
n
t
o
f
Artifi
c
ial
I
n
telli
g
e
n
c
e
,
M
a
c
h
in
e
Lea
rn
in
g
a
n
d
Da
ta
S
c
ien
c
e
a
t
CHRIST
Un
iv
e
rsity
,
Ba
n
g
a
l
o
re
.
He
c
o
m
p
l
e
ted
h
is
P
h
.
D.
d
e
g
re
e
in
C
o
m
p
u
t
e
r
S
c
ien
c
e
a
n
d
En
g
in
e
e
ri
n
g
fro
m
VTU,
Be
lg
a
u
m
,
In
d
ia
i
n
2
0
1
7
a
n
d
h
a
s
m
o
re
th
a
n
2
1
+
y
e
a
rs
o
f
tea
c
h
in
g
e
x
p
e
rien
c
e
.
His
in
tere
sts
in
c
l
u
d
e
d
a
ta
m
in
i
n
g
,
a
r
ti
ficia
l
in
telli
g
e
n
c
e
,
a
n
d
b
ig
d
a
t
a
.
He
c
a
n
b
e
c
o
n
tac
ted
at
e
m
a
il
:
ra
g
h
a
v
.
tr
g
@g
m
a
il
.
c
o
m
.
S
a
n
t
o
sh
K
u
m
a
r
J
a
n
k
a
tti
is
c
u
rre
n
tl
y
wo
rk
in
g
a
s
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
t
h
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
Tec
h
n
o
l
o
g
y
a
t
Da
y
a
n
a
n
d
a
S
a
g
a
r
Un
iv
e
rsity
,
Ba
n
g
a
lo
re
.
He
c
o
m
p
lete
d
h
is
P
h
.
D
.
d
e
g
re
e
i
n
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
fro
m
VTU,
Be
lg
a
u
m
,
In
d
ia
i
n
2
0
2
2
a
n
d
h
a
s
m
o
re
th
a
n
1
4
y
e
a
rs
o
f
tea
c
h
i
n
g
e
x
p
e
rien
c
e
a
n
d
3
y
e
a
rs
o
f
IT
i
n
d
u
stry
e
x
p
e
rien
c
e
.
His
in
tere
sts
in
c
lu
d
e
d
a
ta
m
i
n
in
g
,
a
rti
ficia
l
in
tell
ig
e
n
c
e
,
a
n
d
b
i
g
d
a
ta.
He
c
a
n
b
e
c
o
n
tac
ted
at
e
m
a
il
:
sja
n
k
a
t
ti
@g
m
a
il
.
c
o
m
.
Ma
n
ju
n
a
th
Ra
m
a
n
n
a
La
m
a
n
i
h
o
ld
s
a
P
h
.
D.
i
n
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
i
n
e
e
rin
g
fr
o
m
CHRIST
Un
iv
e
rsity
.
He
i
s
c
u
rre
n
tl
y
wo
rk
in
g
a
s
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
,
M
o
o
d
lak
a
tt
e
In
stit
u
te
o
f
Tec
h
n
o
lo
g
y
,
Ku
n
d
a
p
u
ra
,
Ud
u
p
i
,
Ka
rn
a
tak
a
,
In
d
ia.
His
a
c
a
d
e
m
ic
in
tere
sts
sp
a
n
d
e
e
p
lea
rn
in
g
,
AI,
Io
T
,
a
n
d
p
ro
g
ra
m
m
in
g
.
He
c
a
n
b
e
c
o
n
tac
te
d
at
e
m
a
il
:
m
a
n
ju
n
a
th
lam
a
n
i0
1
@
g
m
a
il
.
c
o
m
.
Nir
a
n
ja
n
a
S
h
r
a
v
a
n
a
b
e
la
g
o
la
J
in
a
c
h
a
n
d
r
a
c
o
m
p
lete
d
h
is
P
h
.
D.
fro
m
VTU,
Be
lag
a
v
i
in
2
0
2
2
.
He
h
a
s
d
o
n
e
h
i
s
m
a
ste
rs
in
M
a
c
h
in
e
De
sig
n
fr
o
m
VTU,
Be
lag
a
v
i.
His
a
re
a
s
o
f
in
tere
st
a
re
ima
g
e
p
ro
c
e
ss
in
g
,
m
a
c
h
in
e
lea
rn
in
g
,
a
n
d
flu
i
d
d
y
n
a
m
ics
.
He
is
c
u
rre
n
tl
y
wo
rk
i
n
g
a
s
a
ss
istan
t
p
ro
fe
ss
o
r
i
n
t
h
e
De
p
a
rtme
n
t
o
f
M
e
c
h
a
n
ica
l
En
g
i
n
e
e
rin
g
a
t
CHRIS
T
Un
iv
e
rsity
.
He
c
a
n
b
e
c
o
n
tac
ted
at
e
m
a
il
:
sjn
iran
jan
8
6
@
g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.