I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
. 14, No. 5, O
c
to
be
r
2025
, pp.
4050
~
4060
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
14
.i
5
.pp
4050
-
4060
4050
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
L
au
r
e
n
t
se
r
i
e
s
i
n
t
e
l
l
i
ge
n
t
m
u
l
t
i
d
i
m
e
n
si
on
al
ob
j
e
c
t
op
t
i
m
i
z
at
i
on
c
l
ass
i
f
i
c
at
i
on
f
or
c
r
op
d
i
se
ase
d
e
t
e
c
t
i
on
A
n
an
d
h
an
K
ar
u
n
a
n
it
h
i
1,
2
,
A
j
ay
S
h
an
k
e
r
S
in
g
h
2
1
D
e
pa
r
t
m
e
nt
of
C
om
put
e
r
S
c
i
e
nc
e
a
nd
E
ngi
ne
e
r
i
ng,
G
I
T
A
M
D
e
e
m
e
d
to
be
U
ni
ve
r
s
i
t
y,
B
e
nga
l
ur
u,
I
ndi
a
2
S
c
hool
of
C
om
put
i
ng
S
c
i
e
nc
e
a
nd
E
ngi
ne
e
r
i
ng,
G
a
l
got
i
a
s
U
ni
ve
r
s
i
t
y,
G
r
e
a
t
e
r
N
oi
da
,
I
ndi
a
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
J
un 6, 2024
R
e
vi
s
e
d
J
un 28, 2025
A
c
c
e
pt
e
d
A
ug 6, 2025
Rice
crop
disease
detection
and
its
diagnosis
methods
are
vitally
im
portant
for
the
agricult
ure
field
to
be
sustainable.
T
raditional
methods
suffe
r
from
paddy
yield,
complex
issues,
and
crop
diseases
,
leading
to
inefficie
n
cies
in
the
agricult
ure
domain.
Our
research
provides
space
for
a
novel
app
roach,
combini
ng
the
Laurent
series
with
an
intelligent
multidimensional
object
optimization
(LIMO)
classifi
cation
framework
based
on
gen
erative
adversarial
networks
(GANs)
to
recognize
various
types
of
crop
dise
ases
in
agricult
ural
fields.
Through
our
proposed
research
work,
IoT
nodes
se
nse
the
values
of
the
field
crop,
and
gathered
information
is
shared
with
proc
essing
units
through
base
station
communi
cation.
M
ulti
-
objective
and
cogniti
ve
learning
routing
(MOCLEAR)
protocol
supports
choosing
the
optim
al
path
for
data
transmission
improvement.
Then,
for
image
segmentation,
GAN
combined
with
cogniti
ve
residual
convolut
ion
network
(CRCN
et)
is
modified
to
segment
values
from
input
images.
After
receiving
se
gment
input
images,
perform
feature
extractio
n
and
classifi
cation
using
sign
ificant
attribut
es.
The
proposed
Laurent
s
eries
with
IMO
is
newly
formula
ted
by
integrating
the
Laurent
series
with
Intelligen
t
IMO
algorit
hms.
T
hrough
extensiv
e
experiment
ation
and
analysis
,
t
he
proposed
LIMO
-
based
GAN
network
provides
effective
and
improved
performance
metrics
with
overall
accuracy,
sensitivity,
and
specificity
values
at
91.5%,
92.6%,
and
9
2.41%,
respectively.
K
e
y
w
o
r
d
s
:
C
r
op dis
e
a
s
e
s
G
e
ne
r
a
ti
ve
a
dve
r
s
a
r
ia
l
ne
twor
ks
I
nt
e
ll
ig
e
nt
m
ul
ti
di
m
e
ns
io
na
l
L
a
ur
e
nt
s
e
r
ie
s
L
I
M
O
c
la
s
s
if
ic
a
ti
on
M
ul
ti
-
obj
e
c
ti
ve
a
nd
c
ogni
ti
ve
le
a
r
ni
ng
r
out
in
g
This
is
an
open
access
article
under
the
CC
BY
-
SA
license.
C
or
r
e
s
pon
di
n
g
A
u
th
or
:
A
na
ndha
n
K
a
r
una
ni
th
i
D
e
pa
r
tm
e
nt
of
C
om
put
e
r
S
c
ie
nc
e
a
nd
E
ngi
ne
e
r
in
g,
G
I
T
A
M
D
e
e
m
e
d
to
be
U
ni
ve
r
s
it
y
B
e
nga
lu
r
u,
K
a
r
na
ta
ka
,
I
ndi
a
E
m
a
il
:
a
na
ndha
npg13
@
gm
a
il
.c
om
1.
I
N
T
R
O
D
U
C
T
I
O
N
R
ic
e
is
a
s
ta
pl
e
f
ood
f
or
a
s
ig
ni
f
ic
a
nt
po
r
ti
on
of
th
e
gl
oba
l
pop
ul
a
ti
on,
a
nd
its
c
ul
ti
va
ti
on
is
vi
ta
l
f
or
f
ood
s
e
c
ur
it
y,
e
s
pe
c
ia
ll
y
in
c
ount
r
ie
s
li
ke
I
ndi
a
,
w
he
r
e
a
gr
ic
ul
tu
r
e
f
or
m
s
th
e
ba
c
kbone
of
th
e
e
c
onomy.
H
ow
e
ve
r
,
r
ic
e
c
r
ops
a
r
e
hi
ghl
y
s
us
c
e
pt
ib
le
to
va
r
io
us
di
s
e
a
s
e
s
,
w
hi
c
h
can
le
a
d
to
s
ubs
ta
nt
ia
ll
y
le
s
s
pr
oduc
ti
vi
ty
,
r
e
duc
e
d
f
ood
s
e
c
ur
it
y,
a
nd
e
c
onomi
c
di
f
f
ic
ul
ty
f
or
f
a
r
m
e
r
s
.
In
e
a
r
li
e
r
m
e
th
ods
of
di
s
e
a
s
e
de
te
c
ti
on
a
r
e
of
te
n
m
a
nu
a
l,
time
-
c
ons
um
in
g,
a
nd
pr
one
to
e
r
r
or
s
,
le
a
di
ng
to
d
e
la
ye
d
in
te
r
ve
nt
io
ns
a
nd
f
ur
th
e
r
c
r
op
da
m
a
ge
[
1]
.
R
ic
e
pl
a
nt
s
a
r
e
vul
ne
r
a
bl
e
to
v
a
r
io
us
le
a
f
di
s
e
a
s
e
s
,
w
hi
c
h
can
s
ig
ni
f
ic
a
nt
ly
im
pa
c
t
c
r
op
yi
e
ld
a
nd
qua
li
ty
.
T
he
s
e
di
s
e
a
s
e
s
,
c
a
u
s
e
d
by
f
ungi
,
ba
c
te
r
ia
,
or
vi
r
us
e
s
,
of
te
n
m
a
ni
f
e
s
t
as
di
s
c
ol
or
a
ti
on,
le
s
io
ns
,
or
w
il
ti
ng,
a
f
f
e
c
ti
ng
th
e
pl
a
nt
’
s
a
bi
li
ty
to
phot
os
ynt
h
e
s
iz
e
e
f
f
ic
ie
nt
ly
.
O
ur
pr
im
a
r
y
f
oc
us
is
on
th
e
de
te
c
ti
on
a
nd
di
a
gno
s
is
of
r
ic
e
c
r
op
di
s
e
a
s
e
s
u
s
in
g
a
dva
n
c
e
d
AI
te
c
hni
que
s
,
s
pe
c
if
ic
a
ll
y
a
nov
e
l
f
r
a
m
e
w
or
k
th
a
t
c
om
bi
ne
s
th
e
L
a
ur
e
nt
s
e
r
ie
s
w
it
h
in
te
ll
ig
e
nt
m
ul
ti
di
m
e
ns
io
na
l
obj
e
c
t
opt
im
iz
a
ti
on
(
L
I
M
O
)
a
nd
ge
ne
r
a
ti
ve
a
dve
r
s
a
r
ia
l
n
e
twor
ks
(
G
A
N
s
)
.
T
hi
s
not
onl
y
h
e
lp
s
in
r
e
duc
in
g
c
r
op
lo
s
s
e
s
but
a
ls
o
s
uppor
ts
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
L
aur
e
nt
s
e
r
i
e
s
i
nt
e
ll
ig
e
nt
m
ul
ti
di
m
e
ns
io
nal
obj
e
c
t
opt
imi
z
at
io
n c
la
s
s
if
ic
at
io
n
…
(
A
nandhan
K
ar
unanit
hi
)
4051
s
us
ta
in
a
bl
e
a
gr
ic
ul
tu
r
a
l
pr
a
c
ti
c
e
s
by
m
in
im
iz
in
g
th
e
ne
e
d
f
or
e
xc
e
s
s
iv
e
p
e
s
ti
c
id
e
u
s
e
[
2]
.
Y
ou
onl
y
lo
ok
onc
e
(
Y
O
L
O
)
pr
e
vi
ous
ve
r
s
io
n
di
d
not
ha
ve
a
c
ons
tr
a
in
t
on
lo
c
a
ti
on
pr
e
di
c
ti
on,
m
a
ki
ng
it
uns
ta
bl
e
on
e
a
r
ly
it
e
r
a
ti
ons
.
T
he
Y
O
L
O
v2
pr
e
di
c
ts
f
iv
e
pa
r
a
m
e
te
r
s
a
nd
a
ppl
ie
s
t
he
da
r
kne
t
f
unc
ti
on
to
a
c
ons
tr
a
in
t
if
its
va
lu
e
is
be
twe
e
n
0
a
nd
1.
T
he
Y
O
L
O
v
2
m
e
th
od
w
a
s
im
pl
e
m
e
nt
e
d
w
it
h
a
hi
dde
n
la
ye
r
f
or
e
f
f
e
c
ti
ve
obj
e
c
t
r
e
tr
ie
va
l
in
th
e
s
m
a
r
t
a
gr
ic
ul
tu
r
e
f
ie
ld
.
Y
O
L
O
v
2
in
d
ic
a
te
s
in
f
or
m
a
ti
on
to
f
a
r
m
in
g
pe
opl
e
a
f
te
r
th
e
id
e
nt
if
ic
a
ti
on
of
r
ic
e
c
r
op
di
s
e
a
s
e
s
.
A
l
s
o,
Y
O
L
O
v
2
im
pr
ove
s
pr
oduc
ti
vi
ty
w
i
th
ne
e
de
d
in
f
or
m
a
ti
on.
T
hi
s
Y
O
L
O
v
2
m
ode
l,
th
r
oug
h
th
e
s
e
ns
or
,
r
e
c
e
iv
e
s
pl
a
nt
le
a
f
di
s
e
a
s
e
im
a
g
e
s
a
nd
is
r
e
f
in
e
d
w
it
h
a
m
e
di
a
n
f
il
te
r
a
f
te
r
th
e
s
e
gm
e
nt
a
ti
on
a
nd
c
la
s
s
if
ic
a
ti
on
pr
oc
e
s
s
[
3]
.
An
a
dva
nc
e
d
opt
im
iz
e
d
a
lg
or
it
hm
s
uppor
ts
th
e
L
I
M
O
m
ode
l
to
t
r
a
in
th
e
r
ic
e
c
r
o
p
di
s
e
a
s
e
da
t
a
s
e
t
to
th
e
m
a
c
hi
ne
.
I
oT
ha
s
pl
e
nt
y
of
oppor
tu
ni
ti
e
s
a
nd
ha
s
c
ont
r
ib
ut
e
d
a
vi
ta
l
r
ol
e
in
w
ir
e
le
s
s
ne
twor
ks
,
e
s
pe
c
ia
ll
y
in
th
e
la
s
t
f
if
te
e
n
ye
a
r
s
.
V
a
r
io
us
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
m
e
th
ods
ne
e
d
m
or
e
a
c
c
ur
a
c
y
a
nd
di
m
e
ns
io
na
l
c
or
r
e
c
ti
ons
.
T
hr
ough
th
is
r
e
s
e
a
r
c
h
w
or
k,
pl
a
nni
ng
to
ge
a
r
up
th
e
a
c
c
ur
a
c
y
of
th
e
r
ic
e
(
pa
ddy)
c
r
op
di
s
e
a
s
e
de
t
e
c
ti
on
s
ys
t
e
m
w
it
h
th
e
e
f
f
e
c
t
of
a
ne
w
-
f
a
ngl
e
d
im
a
ge
c
la
s
s
if
ic
a
ti
on
a
lg
or
it
hm
w
it
h
c
ogni
ti
ve
le
a
r
ni
ng
[
4]
.
T
hr
ough
th
is
r
e
s
e
a
r
c
h
w
or
k,
r
ic
e
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
a
nd
pr
e
ve
nt
io
n
w
it
h
th
e
he
lp
of
G
A
N
ne
twor
ks
,
w
hi
c
h
pr
ovi
de
pa
r
a
ll
e
l
s
ig
ni
f
ic
a
nt
s
pe
e
d
-
up
s
a
m
pl
e
s
(
in
put
im
a
ge
s
)
.
A
ls
o,
GANs
a
r
e
ge
tt
in
g
tr
a
in
da
ta
th
r
ough
L
I
M
O
'
s
pr
opos
e
d
f
r
a
m
e
w
or
k,
w
hi
c
h
in
te
gr
a
te
s
bot
h
th
e
L
I
M
O
a
lg
or
it
hm
to
im
pr
ove
th
e
ove
r
a
ll
hi
t
r
a
te
a
nd
a
c
c
ur
a
c
y
of
r
ic
e
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
a
nd
pr
e
ve
nt
io
n.
F
or
th
e
te
xt
ur
e
c
la
s
s
if
ic
a
ti
on
a
ppr
oa
c
h,
us
in
g
a
gr
a
y
le
ve
l
co
-
oc
c
ur
r
e
nc
e
m
a
tr
ix
th
a
t
c
on
s
id
e
r
s
ov
e
r
la
p
a
nd
e
dg
e
pi
xe
l
e
xt
r
a
c
t
e
s
s
e
nt
ia
l
a
tt
r
ib
ut
e
s
is
r
e
vi
s
e
d
by
th
e
G
A
N
ne
twor
k,
w
hi
c
h
is
tr
a
in
e
d
by
L
I
M
O
.
T
he
ge
ne
r
a
to
r
c
r
e
a
te
s
s
ynt
he
ti
c
im
a
ge
s
of
c
r
op
d
is
e
a
s
e
s
ba
s
e
d
on
th
e
in
put
da
ta
,
w
hi
le
th
e
di
s
c
r
im
in
a
to
r
e
va
lu
a
te
s
th
e
a
ut
he
nt
ic
it
y
of
th
e
s
e
im
a
ge
s
.
T
he
two
n
e
twor
ks
a
r
e
tr
a
in
e
d
s
im
ul
ta
ne
ous
ly
,
w
it
h
th
e
ge
n
e
r
a
to
r
im
pr
ovi
ng
its
a
bi
li
ty
to
c
r
e
a
te
r
e
a
li
s
ti
c
im
a
ge
s
a
nd
th
e
di
s
c
r
im
in
a
to
r
be
c
om
in
g
be
tt
e
r
at
di
s
ti
ngui
s
hi
ng
b
e
twe
e
n
r
e
a
l
a
nd
s
ynt
he
ti
c
im
a
ge
s
.
G
A
N
s
a
r
e
p
a
r
ti
c
ul
a
r
ly
e
f
f
e
c
ti
ve
in
s
c
e
na
r
io
s
w
he
r
e
th
e
a
va
il
a
bl
e
da
ta
s
e
t
is
li
m
it
e
d
or
im
ba
la
nc
e
d
.
By
ge
ne
r
a
ti
ng
s
ynt
he
ti
c
im
a
ge
s
,
G
A
N
s
he
lp
c
r
e
a
te
a
m
or
e
di
ve
r
s
e
a
nd
r
e
pr
e
s
e
nt
a
ti
ve
da
ta
s
e
t,
w
hi
c
h
im
pr
ove
s
th
e
pe
r
f
or
m
a
nc
e
of
th
e
c
la
s
s
if
ic
a
ti
on
m
ode
l.
T
he
pr
opos
e
d
s
ys
te
m
le
ve
r
a
g
e
s
I
oT
te
c
hnol
ogy
to
c
ol
le
c
t
r
e
a
l
-
time
da
ta
f
r
om
a
g
r
ic
ul
tu
r
a
l
f
ie
ld
s
[
5]
,
w
hi
c
h
is
th
e
n
pr
oc
e
s
s
e
d
u
s
in
g
a
m
ul
ti
-
obj
e
c
ti
ve
a
nd
c
ogni
ti
ve
le
a
r
ni
ng
r
out
in
g
(
M
O
C
L
E
A
R
)
pr
ot
oc
ol
f
or
opt
im
a
l
da
ta
tr
a
ns
m
is
s
io
n.
T
h
e
f
r
a
m
e
w
or
k
a
im
s
to
im
pr
ove
th
e
a
c
c
ur
a
c
y
a
nd
e
f
f
ic
ie
nc
y
of
r
ic
e
c
r
op
di
s
e
a
s
e
de
te
c
ti
on,
w
hi
c
h
is
c
r
it
ic
a
l
f
or
s
us
ta
in
a
bl
e
a
gr
ic
ul
tu
r
e
[
6]
.
T
hi
s
pa
pe
r
is
w
or
th
r
e
a
di
ng
be
c
a
us
e
it
in
tr
oduc
e
s
a
nove
l
a
ppr
oa
c
h
th
a
t
c
om
bi
ne
s
th
e
L
a
ur
e
nt
s
e
r
ie
s
w
it
h
L
I
M
O
a
nd
G
A
N
s
to
c
r
e
a
te
a
m
or
e
a
c
c
ur
a
te
a
nd
e
f
f
ic
ie
nt
s
ys
te
m
f
or
r
ic
e
c
r
op
di
s
e
a
s
e
de
te
c
ti
on.
T
he
pr
opos
e
d
f
r
a
m
e
w
or
k
a
ddr
e
s
s
e
s
th
e
li
m
it
a
ti
ons
of
e
xi
s
ti
ng
m
e
th
ods
by
in
te
gr
a
ti
ng
a
dva
nc
e
d
AI
te
c
hni
que
s
w
it
h
I
oT
,
e
na
bl
in
g
r
e
a
l
-
time
m
oni
to
r
in
g
a
nd
e
a
r
ly
de
te
c
ti
on
of
di
s
e
a
s
e
s
[
7]
.
T
he
r
e
s
e
a
r
c
h'
s
m
a
in
a
im
is
to
im
pl
e
m
e
nt
a
f
our
-
pha
s
e
f
r
a
m
e
w
or
k
f
or
r
ic
e
c
r
op
di
s
e
a
s
e
de
te
c
ti
on:
I
oT
c
om
m
uni
c
a
ti
on
pha
s
e
:
s
e
ns
or
s
c
ol
le
c
t
da
ta
f
r
om
th
e
f
ie
ld
,
w
hi
c
h
is
tr
a
ns
m
it
te
d
to
a
pr
oc
e
s
s
in
g
uni
t
us
in
g
th
e
M
O
C
L
E
A
R
pr
ot
oc
ol
f
or
opt
im
a
l
r
out
in
g.
P
r
e
-
pr
oc
e
s
s
in
g
ph
a
s
e
:
T
he
c
ol
le
c
te
d
da
ta
und
e
r
goe
s
d
a
ta
r
e
duc
ti
on
a
nd
f
e
a
tu
r
e
e
ngi
ne
e
r
in
g
to
im
pr
ove
th
e
qua
li
ty
of
in
put
va
lu
e
s
f
or
s
e
gm
e
nt
a
ti
on
a
nd
c
la
s
s
if
ic
a
ti
on.
I
m
a
ge
pr
oc
e
s
s
in
g
pha
s
e
:
th
e
c
ogni
ti
ve
r
e
s
id
ua
l
c
onvolut
io
n
ne
twor
k
(
C
R
C
N
e
t
)
is
us
e
d
f
or
im
a
ge
s
e
gm
e
nt
a
ti
on,
f
ol
lo
w
e
d
by
f
e
a
tu
r
e
e
xt
r
a
c
ti
on
a
nd
c
la
s
s
if
ic
a
ti
on
u
s
in
g
th
e
L
I
M
O
f
r
a
m
e
w
or
k.
D
is
e
a
s
e
de
te
c
ti
on
ph
a
s
e
:
th
e
L
I
M
O
f
r
a
m
e
w
or
k,
c
om
bi
ne
d
w
it
h
G
A
N
s
,
c
la
s
s
if
ie
s
th
e
di
s
e
a
s
e
a
nd
pr
ovi
de
s
in
s
ig
ht
s
in
to
th
e
s
e
v
e
r
it
y
of
th
e
in
f
e
c
ti
on.
T
he
pa
pe
r
a
ls
o
in
c
lu
de
s
a
c
om
p
a
r
a
ti
ve
a
na
ly
s
is
of
t
he
pr
opos
e
d
f
r
a
m
e
w
or
k
w
it
h
e
xi
s
ti
ng
m
ode
ls
,
de
m
ons
tr
a
ti
ng
its
s
upe
r
io
r
pe
r
f
o
r
m
a
nc
e
in
te
r
m
s
of
a
c
c
ur
a
c
y,
s
e
ns
it
iv
it
y,
a
nd
s
pe
c
if
ic
it
y.
T
he
s
tr
uc
tu
r
e
of
th
e
s
e
c
ti
ons
is
as
f
ol
lo
w
s
:
s
e
c
ti
on
2
pr
ovi
de
s
a
r
e
vi
e
w
of
r
e
la
te
d
w
or
k
in
th
e
f
ie
ld
of
c
r
op
di
s
e
a
s
e
de
te
c
ti
on,
hi
ghl
ig
ht
in
g
th
e
li
m
it
a
ti
ons
of
e
xi
s
ti
ng
m
e
th
ods
.
S
e
c
ti
on
3
de
s
c
r
ib
e
s
th
e
m
a
te
r
ia
ls
a
nd
te
c
hni
que
s
us
e
d
in
th
e
pr
opos
e
d
f
r
a
m
e
w
or
k,
in
c
lu
di
ng
th
e
da
ta
s
e
t,
I
oT
c
om
m
uni
c
a
ti
on,
a
nd
th
e
L
I
M
O
c
la
s
s
if
ic
a
ti
on
f
r
a
m
e
w
or
k.
S
e
c
ti
on
4
de
ta
il
s
th
e
e
xpe
r
im
e
nt
a
l
r
e
s
ul
ts
,
c
om
pa
r
in
g
th
e
p
e
r
f
or
m
a
nc
e
of
th
e
pr
opos
e
d
f
r
a
m
e
w
or
k
w
it
h
e
xi
s
ti
ng
m
ode
ls
.
S
e
c
ti
on
5
c
onc
lu
de
s
th
e
pa
pe
r
,
s
um
m
a
r
iz
i
ng
th
e
ke
y
f
in
di
ngs
a
nd
s
ugge
s
ti
ng
f
ut
ur
e
di
r
e
c
ti
ons
f
or
r
e
s
e
a
r
c
h.
2.
R
E
L
A
T
E
D
WO
R
K
E
a
r
li
e
r
in
ve
s
ti
ga
ti
ons
in
th
e
dom
a
in
of
r
ic
e
di
s
e
a
s
e
c
la
s
s
if
ic
a
ti
on
us
in
g
de
e
p
le
a
r
ni
ng
m
e
th
ods
ha
ve
be
e
n
di
r
e
c
te
d
e
xt
e
ns
iv
e
ly
by
e
xpe
r
ts
gl
oba
ll
y.
T
hi
s
s
tu
dy
i
nve
s
ti
ga
te
d
th
e
e
f
f
e
c
ts
of
I
oT
,
an
im
por
ta
nt
te
c
hnol
ogy
th
a
t
ne
e
ds
to
c
onne
c
t
c
om
m
uni
c
a
ti
on
de
vi
c
e
s
to
th
e
in
te
r
ne
t
.
It
m
a
ke
s
an
im
m
e
ns
e
s
m
a
r
t
a
ppl
ic
a
ti
on.
T
h
e
I
oT
ha
s
gr
ow
n
r
a
pi
dl
y,
a
nd
th
a
t
is
th
e
r
e
a
s
on
m
a
ny
r
e
s
e
a
r
c
h
e
r
s
a
r
e
s
how
in
g
in
te
r
e
s
t
in
th
is
pl
a
tf
or
m
.
An
a
gr
o
-
w
e
a
th
e
r
s
ta
ti
on
in
s
ta
ll
e
d
in
an
a
gr
ic
ul
tu
r
e
f
ie
ld
,
th
r
ough
th
is
pl
a
tf
or
m
,
can
c
ol
le
c
t
ve
r
s
a
ti
le
in
f
or
m
a
ti
on
w
it
h
va
r
io
us
s
e
ns
or
s
f
or
th
e
a
ppl
ic
a
ti
on'
s
s
us
ta
in
a
bi
li
ty
.
V
a
r
io
us
di
m
e
ns
io
na
l
f
e
a
tu
r
e
s
a
r
e
e
xt
r
a
c
ts
a
nd
de
te
c
ti
ng
c
r
op
di
s
e
a
s
e
s
in
an
e
f
f
e
c
ti
ve
m
a
nne
r
.
W
it
h
th
i
s
c
om
put
a
ti
ona
l
in
te
ll
ig
e
n
c
e
,
we
can
a
bl
e
to
pe
r
f
or
m
num
e
r
ous
s
e
ns
or
s
a
nd
c
om
m
uni
c
a
ti
on
(
ne
twor
k)
de
vi
c
e
s
f
or
s
m
a
r
t
a
gr
ic
ul
tu
r
a
l
pr
oc
e
s
s
e
s
,
w
hi
c
h
im
pr
ove
e
xt
r
e
m
e
pr
oduc
ti
vi
ty
.
C
lo
ud
m
ode
ls
,
m
a
c
hi
ne
le
a
r
ni
ng,
a
nd
I
oT
m
a
te
r
ia
li
z
a
ti
on
te
c
hnol
ogi
e
s
pl
a
y
a
vi
ta
l
r
ol
e
in
f
ie
ld
in
f
or
m
a
ti
on
c
ol
le
c
ti
on
f
or
s
m
a
r
t
f
a
r
m
in
g
[
8]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 5, O
c
to
be
r
2025
:
4050
-
4060
4052
T
he
a
ut
ho
r
e
xa
m
in
e
d
ho
w
c
r
op
d
is
e
a
s
e
de
te
c
ti
on
m
e
th
ods
e
f
f
e
c
ti
v
e
l
y
c
on
tr
ol
f
a
r
m
in
g
f
a
c
to
r
s
by
us
in
g
va
r
i
ous
s
e
ns
or
s
w
it
h
th
os
e
va
l
ue
s
,
c
r
e
a
ti
ng
a
c
o
nvo
lu
t
io
na
l
ne
ur
a
l
ne
two
r
k
m
o
de
l
-
t
r
a
in
e
d
da
ta
s
e
t
.
T
hi
s
m
ode
l
a
m
e
nde
d
t
he
ove
r
a
l
l
f
r
a
m
e
w
o
r
k
a
n
d
is
an
e
f
f
e
c
ti
ve
m
on
i
to
r
i
ng
s
ys
te
m
b
a
s
e
d
on
th
e
r
m
a
l
c
a
m
e
r
a
i
m
a
ge
s
to
f
i
nd
w
he
t
he
r
c
r
op
le
a
v
e
s
a
r
e
in
f
e
c
te
d
or
n
ot
.
I
nve
nt
e
d
a
ne
w
f
r
a
m
e
w
o
r
k
f
or
th
e
a
gr
ic
u
lt
ur
a
l
i
nd
us
t
r
y
t
ha
t
p
e
r
f
o
r
m
s
e
dge
c
om
p
ut
in
g
f
o
r
da
ta
t
r
a
ns
m
is
s
io
n
f
r
om
I
oT
s
e
ns
o
r
s
to
c
lo
ud
s
to
r
a
ge
.
T
he
m
a
i
n
ob
je
c
ti
v
e
of
t
hi
s
m
ode
l
is
f
ie
l
d
m
o
ni
to
r
i
n
g,
e
va
lu
a
ti
ng,
a
nd
t
r
a
ns
m
i
tt
in
g
va
l
ue
s
to
t
he
c
lo
u
d.
O
pt
im
iz
a
ti
on
a
ls
o
pl
a
ys
a
v
it
a
l
r
ol
e
in
t
hi
s
m
od
e
l
,
w
hi
c
h
m
e
a
ns
c
o
nt
in
u
ous
m
on
it
o
r
in
g
a
n
d
e
va
l
ua
t
in
g
va
l
ue
s
c
o
m
m
uni
c
a
te
s
th
e
c
l
ou
d
in
a
s
s
o
r
te
d
da
i
r
y
c
i
r
c
um
s
ta
nc
e
s
.
T
he
a
u
th
o
r
va
li
da
t
e
d
t
hi
s
in
f
i
ve
c
r
o
ps
,
17
d
is
e
a
s
e
s
,
a
n
d
a
1
21
,
95
5
i
m
a
g
e
s
da
ta
s
e
t
ov
e
r
f
ie
ld
c
on
di
ti
ons
.
T
he
a
ut
ho
r
i
m
p
le
m
e
nt
e
d
a
de
e
p
le
a
r
ni
n
g
m
o
de
l
a
n
d
a
c
hi
e
ve
d
a
m
u
lt
i
-
c
r
op
c
o
nv
ol
u
ti
o
n
ne
u
r
a
l
ne
two
r
k
(
C
N
N
)
a
c
c
u
r
a
c
y
of
9
8%
.
H
ow
e
ve
r
,
th
is
m
ode
l
is
a
bs
e
nt
in
t
r
a
in
in
g
dyna
m
ic
a
l
ly
th
e
le
a
r
ni
ng
a
l
go
r
i
th
m
s
f
or
m
is
c
e
l
la
ne
ous
c
r
o
p
di
s
e
a
s
e
s
a
n
d
s
ym
pt
om
s
[
8
]
,
[
9]
.
T
hi
s
s
t
udy
in
v
e
s
t
ig
a
te
d
c
r
o
p
di
s
e
a
s
e
d
e
te
c
ti
on
w
it
h
e
no
ugh
w
e
ig
ht
y
li
te
r
a
tu
r
e
r
e
v
ie
w
a
nd
v
a
r
io
us
a
c
t
iv
it
ie
s
to
to
le
r
a
t
e
th
e
ne
w
I
o
T
m
ode
l
f
or
f
a
r
m
in
g.
O
n
e
m
o
r
e
th
in
g
is
th
e
I
oT
p
la
t
f
o
r
m
pr
ov
id
e
s
us
e
r
-
f
r
ie
n
dl
y
m
od
e
ls
f
o
r
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
a
n
d
pr
e
ve
n
ti
on
.
T
he
I
oT
m
ode
l
di
r
e
c
t
ly
c
o
nne
c
ts
f
a
r
m
e
r
s
w
i
th
c
o
m
p
ut
a
ti
on
(
in
te
l
li
ge
n
c
e
c
o
m
pu
ta
t
io
n
)
.
In
t
hi
s
s
t
udy
,
th
e
a
u
th
o
r
a
c
h
ie
v
e
d
an
a
c
c
ur
a
c
y
of
82
.41
%
,
a
n
d
t
he
F1
-
s
c
o
r
e
of
67
%
w
a
s
t
he
lo
w
e
s
t
a
m
on
g
a
ll
th
e
c
la
s
s
if
ie
r
s
.
T
he
s
m
a
r
t
a
gr
ic
u
lt
ur
e
s
ys
t
e
m
w
o
r
ks
f
a
s
t
a
nd
a
c
c
u
r
a
t
e
ly
on
c
r
op
d
is
e
a
s
e
i
de
nt
i
f
ic
a
t
io
n
w
i
th
s
u
m
m
a
r
is
e
d
di
f
f
e
r
e
n
t
t
r
a
i
n
s
e
ts
f
o
r
m
e
t
ic
u
lo
us
ne
s
s
in
th
e
a
g
r
ic
u
lt
ur
e
f
i
e
ld
[
1
0]
.
F
r
o
m
th
e
a
b
ove
li
te
r
a
t
ur
e
s
u
r
ve
y
,
di
f
f
e
r
e
nt
s
ol
u
ti
ons
f
o
r
pl
a
nt
d
is
e
a
s
e
de
t
e
c
t
io
n
m
e
t
hods
a
r
e
not
up
to
e
xpe
c
ta
ti
ons
.
3.
M
E
T
H
O
D
3.1
.
E
xp
e
r
im
e
n
t
al
s
e
t
u
p
T
he
e
xpe
r
im
e
nt
w
a
s
c
ondu
c
te
d
us
in
g
M
A
T
L
A
B
R
2023b,
le
v
e
r
a
gi
ng
its
r
obus
t
im
a
ge
pr
oc
e
s
s
in
g,
de
e
p
le
a
r
ni
ng
to
ol
boxe
s
f
or
s
e
gm
e
nt
a
ti
on
a
nd
c
la
s
s
if
ic
a
ti
on
ta
s
ks
.
T
he
s
ys
te
m
w
a
s
c
onf
ig
ur
e
d
w
it
h
a
W
in
dow
s
11
op
e
r
a
ti
ng
e
nvi
r
onm
e
nt
,
a
64
-
bi
t
a
r
c
hi
te
c
tu
r
e
,
a
n
d
16
GB
R
A
M
.
B
ot
h
th
e
tr
a
in
in
g
a
nd
te
s
ti
ng
pr
oc
e
s
s
of
L
I
M
O
w
it
h
G
A
N
is
f
a
c
il
it
a
te
d
w
it
h
T
e
ns
or
F
lo
w
2.1
5
as
w
e
ll
as
K
e
r
a
s
2.x.
T
he
L
a
ur
e
nt
s
e
r
ie
s
is
a
m
a
th
e
m
a
ti
c
a
l
to
ol
us
e
d
in
c
om
pl
e
x
a
na
ly
s
is
to
r
e
pr
e
s
e
nt
f
unc
ti
ons
w
it
h
s
in
gul
a
r
it
ie
s
.
In
our
f
r
a
m
e
w
or
k,
we
in
te
gr
a
te
th
e
L
I
M
O
to
e
nha
nc
e
th
e
a
c
c
ur
a
c
y
of
c
r
op
di
s
e
a
s
e
de
t
e
c
ti
on.
3.2
.
D
at
a
ac
q
u
is
it
io
n
N
ow
a
da
ys
,
we
a
r
e
e
qui
ppe
d
w
it
h
di
f
f
e
r
e
nt
ty
pe
s
of
s
e
ns
or
s
in
f
a
r
m
in
g
us
in
g
th
e
s
e
ns
e
d
da
ta
u
s
e
d
f
or
our
r
e
s
e
a
r
c
h
w
or
k.
T
he
T
a
bl
e
1
li
s
te
d
ov
e
r
vi
e
w
of
di
s
e
a
s
e
s
w
it
h
a
num
be
r
of
im
a
ge
s
a
nd
r
e
s
ol
ut
io
ns
.
T
he
r
ic
e
c
r
op
di
s
e
a
s
e
da
ta
s
e
t
c
ont
a
in
s
va
r
io
us
r
ic
e
c
r
op
di
s
e
a
s
e
im
a
ge
s
,
a
r
ound
2
,
400
va
r
io
us
c
r
op
di
s
e
a
s
e
im
a
ge
s
(
va
r
io
us
r
a
nge
of
r
e
s
ol
ut
io
n
jp
e
g
f
or
m
a
t
im
a
ge
s
)
da
ta
s
e
ts
us
e
d
f
or
th
is
e
xpe
r
im
e
nt
.
T
he
di
m
e
ns
io
na
li
ty
of
each
da
ta
s
e
t
w
a
s
v
a
r
io
us
qua
nt
it
ie
s
,
va
r
io
us
s
i
z
e
s
of
im
a
ge
s
,
a
n
d
th
e
c
ha
r
a
c
te
r
is
ti
c
of
th
e
a
tt
r
ib
ut
e
(
li
ke
H
S
V
,
gr
a
di
e
nt
,
a
nd
R
G
B
)
.
T
hi
s
da
ta
s
e
t
is
us
e
d
in
th
r
e
e
s
ta
ge
s
:
f
ir
s
t
pr
e
-
pr
oc
e
s
s
in
g,
s
e
c
ond
im
a
ge
e
nha
nc
e
m
e
nt
,
a
nd
s
e
gm
e
nt
a
ti
on,
a
nd
la
s
t
one
is
c
la
s
s
if
ic
a
ti
on
m
e
th
odol
ogy.
T
a
bl
e
1.
D
a
ta
s
e
t
d
e
s
c
r
ip
ti
on
[
11]
–
[
13]
D
i
s
e
a
s
e
t
ype
N
um
be
r
of
i
m
a
ge
s
I
m
a
ge
r
e
s
ol
ut
i
on
B
a
c
t
e
r
i
a
l
bl
i
ght
300
1920
×
1080
B
a
c
t
e
r
i
a
l
l
e
a
f
s
t
r
e
a
k
250
1920
×
1080
B
l
a
s
t
(
l
e
a
f
a
nd
c
ol
l
a
r
)
400
1920
×
1080
F
a
l
s
e
s
m
ut
200
1920
×
1080
R
i
c
e
gr
a
s
s
y
s
t
unt
150
1920
×
1080
B
r
ow
n
s
pot
300
1920
×
1080
T
ot
a
l
i
m
a
ge
s
2
,
400
NA
T
hi
s
a
dva
n
c
e
d
f
r
a
m
e
w
or
k
pr
ovi
de
s
m
a
r
ka
bl
e
im
pr
ove
d
pe
r
f
or
m
a
nc
e
on
c
r
op
di
s
e
a
s
e
d
e
te
c
ti
on
s
ys
te
m
s
,
but
th
e
m
a
in
dr
a
w
ba
c
k
of
th
is
s
ys
te
m
is
its
s
tr
uggl
e
w
it
h
in
c
lu
di
ng
va
r
io
us
ty
pe
s
of
di
s
e
a
s
e
tr
a
in
in
g
da
ta
.
S
tr
e
a
m
in
g
im
a
ge
in
put
s
-
ba
s
e
d
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
s
ys
t
e
m
to
de
te
c
t
th
e
c
r
op
d
is
e
a
s
e
s
a
nd
th
e
ir
le
ve
l
w
a
s
f
or
m
ul
a
te
d
w
it
h
r
e
a
l
-
time
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
[
14]
,
[
15]
.
At
th
e
in
it
ia
l
le
ve
l,
th
e
s
tr
e
a
m
in
g
im
a
ge
w
a
s
c
onve
r
te
d
in
to
s
ta
ti
c
-
m
a
r
gi
ne
d
m
ul
ti
pl
e
im
a
ge
s
.
In
th
e
s
e
c
ond
s
te
p
s
tr
e
a
m
e
d
m
ul
ti
pl
e
im
a
ge
s
w
e
r
e
di
a
gno
s
e
d
w
it
h
obj
e
c
t
id
e
nt
if
ic
a
ti
on.
H
e
r
e
,
th
is
p
r
oc
e
s
s
a
ls
o
ta
ke
s
gr
a
di
e
nt
va
lu
e
s
a
nd
pr
ovi
de
s
m
a
ppi
ng,
w
hi
c
h
is
c
a
ll
e
d
im
a
ge
s
e
gm
e
nt
a
ti
on.
At
th
e
la
s
t,
im
a
ge
s
a
r
e
pr
e
pa
r
e
d
f
or
di
s
e
a
s
e
de
te
c
ti
on
r
e
s
ul
t
s
.
F
ig
ur
e
1
s
how
s
th
e
ov
e
r
a
ll
a
r
c
hi
te
c
tu
r
e
of
c
r
op
di
s
e
a
s
e
de
te
c
ti
on.
T
hi
s
m
ode
l
e
nc
om
pa
s
s
e
s
a
f
e
w
le
ve
ls
onl
y
th
e
r
e
a
r
e
r
a
w
im
a
ge
s
c
ol
le
c
te
d
f
r
om
th
e
a
gr
ic
ul
tu
r
a
l
f
ie
ld
us
in
g
va
r
io
us
s
e
n
s
or
s
,
p
r
e
-
pr
oc
e
s
s
in
g
(
252
×
252
s
iz
e
of
im
a
ge
s
,
da
t
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
L
aur
e
nt
s
e
r
i
e
s
i
nt
e
ll
ig
e
nt
m
ul
ti
di
m
e
ns
io
nal
obj
e
c
t
opt
imi
z
at
io
n c
la
s
s
if
ic
at
io
n
…
(
A
nandhan
K
ar
unanit
hi
)
4053
tr
a
ns
f
or
m
a
ti
on,
da
ta
r
e
duc
ti
on,
a
nd
f
e
a
tu
r
e
e
ngi
ne
e
r
in
g)
,
s
e
gm
e
nt
a
ti
on
us
in
g
C
R
C
N
e
t
,
f
e
a
tu
r
e
le
a
r
ni
ng,
f
e
a
tu
r
e
’
s
e
xt
r
a
c
ti
on
,
c
la
s
s
if
ic
a
ti
on
us
in
g
pr
opos
e
d
L
I
M
O
w
i
th
G
A
N
,
a
nd
e
va
lu
a
ti
on
m
e
tr
ic
s
.
T
he
pl
a
nt
di
s
e
a
s
e
pr
oc
e
s
s
ha
s
to
f
in
d
th
e
de
pt
h
of
di
s
e
a
s
e
s
a
tt
a
c
ki
n
g
a
pa
r
ti
c
ul
a
r
c
r
op
[
16]
–
[
18]
.
T
hi
s
s
e
ve
r
it
y
id
e
nt
if
ic
a
ti
on
is
a
ls
o
di
vi
de
d
in
to
th
r
e
e
pha
s
e
s
li
ke
ne
twor
k
s
(
a
ut
he
nt
ic
,
m
a
li
c
io
us
,
a
nd
a
tt
a
c
ke
r
)
.
T
he
L
a
ur
e
nt
s
e
r
ie
s
a
ll
ow
s
us
to
m
ode
l
c
om
pl
e
x
pa
tt
e
r
ns
in
th
e
da
ta
,
pa
r
ti
c
ul
a
r
ly
in
th
e
pr
e
s
e
nc
e
of
noi
s
e
or
ir
r
e
gul
a
r
i
ti
e
s
in
th
e
im
a
ge
s
.
By
in
c
or
por
a
ti
ng
th
e
L
a
ur
e
nt
s
e
r
ie
s
,
we
can
be
tt
e
r
c
a
pt
ur
e
th
e
unde
r
s
ta
te
d
va
r
ia
ti
ons
in
c
r
op
di
s
e
a
s
e
s
ym
pt
om
s
,
w
hi
c
h
a
r
e
of
te
n
m
is
s
e
d
by
t
r
a
di
ti
ona
l
m
e
th
ods
.
L
I
M
O
is
an
opt
im
iz
a
ti
on
a
lg
or
it
h
m
in
s
pi
r
e
d
by
th
e
hunt
in
g
be
ha
vi
or
of
gr
e
y
w
ol
ve
s
.
It
is
u
s
e
d
to
o
pt
im
iz
e
th
e
f
e
a
tu
r
e
s
e
l
e
c
ti
on
a
nd
c
l
a
s
s
if
ic
a
ti
on
pr
oc
e
s
s
e
s
in
our
f
r
a
m
e
w
or
k.
L
I
M
O
c
om
bi
ne
s
th
e
s
tr
e
ngt
hs
of
I
M
O
w
it
h
c
ogni
ti
ve
le
a
r
ni
ng
te
c
hni
que
s
to
im
pr
ove
th
e
a
c
c
ur
a
c
y
a
nd
e
f
f
ic
ie
nc
y
of
di
s
e
a
s
e
de
te
c
ti
on.
G
A
N
s
a
r
e
a
c
la
s
s
of
d
e
e
p
l
e
a
r
ni
ng
m
ode
ls
th
a
t
c
ons
is
t
of
two
n
e
ur
a
l
ne
twor
ks
:
a
ge
ne
r
a
to
r
a
nd
a
di
s
c
r
im
in
a
t
or
[
19]
.
In
our
f
r
a
m
e
w
or
k,
G
A
N
s
a
r
e
us
e
d
to
ge
ne
r
a
te
s
ynt
h
e
ti
c
im
a
ge
s
of
c
r
op
di
s
e
a
s
e
s
,
w
hi
c
h
a
r
e
th
e
n
u
s
e
d
to
a
ugm
e
nt
th
e
tr
a
in
in
g
da
ta
s
e
t.
T
hi
s
h
e
lp
s
in
im
pr
ovi
ng
th
e
r
obus
tn
e
s
s
a
nd
ge
ne
r
a
li
z
a
ti
on
of
th
e
m
ode
l.
C
R
C
N
e
t
is
a
m
odi
f
ie
d
ve
r
s
io
n
of
th
e
tr
a
di
ti
ona
l
C
N
N
th
a
t
in
c
or
por
a
te
s
c
ogni
ti
ve
le
a
r
ni
ng
a
nd
r
e
s
id
u
a
l
c
onne
c
ti
ons
.
It
is
us
e
d
f
or
im
a
ge
s
e
gm
e
nt
a
ti
on
a
nd
f
e
a
tu
r
e
e
xt
r
a
c
ti
on
in
our
f
r
a
m
e
w
or
k.
M
O
C
L
E
A
R
is
a
r
out
in
g
pr
ot
oc
ol
us
e
d
in
th
e
I
oT
c
om
m
uni
c
a
ti
on
pha
s
e
of
our
f
r
a
m
e
w
or
k.
It
is
de
s
ig
ne
d
to
opt
im
iz
e
th
e
tr
a
ns
m
is
s
io
n
of
da
ta
f
r
om
th
e
s
e
ns
or
s
to
th
e
pr
oc
e
s
s
in
g
uni
t,
e
ns
ur
in
g
m
in
im
a
l
lo
s
s
a
nd
m
a
xi
m
um
e
f
f
ic
ie
nc
y.
B
e
f
or
e
th
e
im
a
ge
s
a
r
e
pr
oc
e
s
s
e
d
by
th
e
C
R
C
N
e
t
a
nd
L
I
M
O
f
r
a
m
e
w
or
ks
,
th
e
y
unde
r
go
a
s
e
r
ie
s
of
pr
e
pr
oc
e
s
s
in
g
s
te
p
s
,
in
c
lu
di
ng
f
e
a
tu
r
e
e
ngi
ne
e
r
in
g
a
nd
da
ta
r
e
duc
ti
on.
F
ig
ur
e
1.
A
r
c
hi
te
c
tu
r
e
of
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
u
s
in
g
L
I
M
O
w
it
h
G
A
N
a
ppr
oa
c
h
m
ode
l
T
he
s
e
s
te
p
s
a
r
e
de
s
ig
n
e
d
to
im
pr
ove
th
e
qua
li
ty
of
th
e
in
put
da
ta
a
nd
r
e
duc
e
th
e
c
om
put
a
ti
ona
l
c
om
pl
e
xi
ty
of
th
e
m
ode
l.
In
th
e
I
oT
c
om
m
uni
c
a
ti
on
pha
s
e
,
s
e
ns
or
s
a
nd
ot
he
r
de
vi
c
e
s
a
r
e
s
ubj
e
c
te
d
to
ne
twor
k
r
out
in
g
p
r
ot
oc
ol
s
.
In
th
is
ne
twor
k
c
om
m
uni
c
a
ti
on
p
ha
s
e
,
pa
th
s
e
le
c
ti
on
a
nd
de
c
i
s
io
n
-
m
a
ki
ng
a
r
e
c
a
r
r
ie
d
out
by
m
ul
ti
-
obj
e
c
ti
ve
a
nd
c
ogni
ti
ve
le
a
r
ni
ng
-
ba
s
e
d
r
out
in
g
a
lg
or
it
hm
s
[
20]
.
O
nc
e
c
om
m
uni
c
a
ti
on
s
ta
r
te
d
w
it
h
M
O
C
L
E
A
R
r
out
in
g
on
boa
r
d
th
e
uni
t
got
in
f
or
m
a
ti
on
a
bout
c
r
op
(
im
a
ge
s
)
.
A
f
te
r
r
e
c
e
iv
in
g
im
a
ge
s
of
th
e
r
ic
e
pl
a
nt
,
r
e
gul
a
r
im
a
ge
pr
oc
e
s
s
in
g
is
pe
r
f
or
m
e
d.
F
e
a
tu
r
e
e
ngi
ne
e
r
in
g
(
pr
e
-
pr
oc
e
s
s
in
g)
s
ta
r
ts
th
e
pr
oc
e
s
s
of
e
xpe
r
t
knowle
dge
f
e
d
to
th
e
m
a
c
hi
ne
.
S
e
gm
e
nt
a
ti
on
a
nd
f
e
a
tu
r
e
e
xt
r
a
c
ti
on
m
e
th
ods
a
r
e
us
e
d
to
a
tt
a
in
in
put
im
a
ge
s
th
r
ough
s
e
ns
or
s
a
nd
I
oT
de
vi
c
e
s
.
F
e
a
tu
r
e
e
ngi
ne
e
r
in
g
to
im
pr
ove
th
e
im
a
ge
qua
li
ty
a
nd
tr
a
ns
f
or
m
a
ti
on
f
or
c
om
f
or
ta
bi
li
ty
.
A
f
te
r
th
a
t,
im
a
ge
s
e
gm
e
nt
a
ti
on
is
c
a
r
r
ie
d
out
ba
s
e
d
on
th
e
e
xpe
c
te
d
obj
e
c
t,
de
s
ig
n,
a
nd
c
ol
or
,
w
hi
c
h
a
r
e
s
e
gm
e
nt
e
d
w
it
h
C
R
C
N
e
t
s
uppor
t
.
F
e
a
tu
r
e
e
xt
r
a
c
ti
on
is
a
c
hi
e
ve
d
us
in
g
p
a
tt
e
r
n,
te
xt
ur
e
,
a
nd
gr
a
di
e
nt
to
in
ha
le
th
e
r
e
qui
r
e
d
f
e
a
tu
r
e
s
to
de
te
c
t
c
r
op
di
s
e
a
s
e
s
[
21]
,
[
22]
.
F
in
a
ll
y,
r
e
c
e
iv
e
d
f
e
a
tu
r
e
s
a
r
e
c
la
s
s
if
ie
d
a
nd
opt
im
iz
e
d,
w
he
r
e
a
s
r
ic
e
pl
a
nt
di
s
e
a
s
e
de
te
c
ti
on
is
done
us
in
g
th
e
L
I
M
O
f
r
a
m
e
w
or
k.
T
he
pr
opos
e
d
L
I
M
O
f
r
a
m
e
w
or
k
m
ode
l
is
a
c
om
bi
na
ti
on
of
L
I
M
O
.
L
I
M
O
f
r
a
m
e
w
or
k
c
om
bi
ne
d
w
it
h
C
R
C
N
e
t
ne
twor
k
f
or
r
ic
e
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
f
or
s
m
a
r
t
a
gr
ic
ul
tu
r
e
w
it
h
I
oT
ne
twor
k.
We
w
il
l
di
s
c
us
s
m
or
e
c
le
a
r
ly
th
e
p
r
oc
e
s
s
of
pa
th
s
e
le
c
ti
on
(
r
out
in
g
a
lg
or
it
h
m
)
a
nd
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
w
it
h
c
ogni
ti
ve
le
a
r
ni
ng.
A
s
s
um
e
r
ic
e
c
r
op
di
s
e
a
s
e
da
ta
s
e
t
R
pd
a
nd
c
ount
of
in
put
va
lu
e
s
.
T
h
e
e
qua
ti
on
is
w
r
it
te
n
as
in
(
1)
.
=
{
1
,
2
,
3
.
.
.
.
.
.
,
}
(
1)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 5, O
c
to
be
r
2025
:
4050
-
4060
4054
W
he
r
e
,
in
di
c
a
te
s
ℎ
pl
a
nt
im
a
ge
s
(
in
put
)
.
A
s
s
um
e
th
a
t
ℎ
node
or
de
vi
c
e
hol
ds
th
e
va
lu
e
s
pe
r
ta
in
in
g
to
s
ubj
e
c
t
f
ie
ld
pl
a
nt
s
or
c
r
op
to
I
oT
c
om
m
uni
c
a
ti
on
pha
s
e
.
R
ic
e
pl
a
nt
da
ta
is
tr
a
ns
m
it
te
d
to
an
onboa
r
d
uni
t
(
B
S
)
,
a
nd
th
e
M
O
C
L
E
A
R
a
lg
or
it
hm
he
lp
s
to
f
in
d
th
e
be
s
t
r
out
e
f
or
da
ta
c
om
m
uni
c
a
ti
on.
3.3
.
C
ogn
it
iv
e
le
ar
n
in
g
-
b
as
e
d
M
O
C
L
E
A
R
al
gor
it
h
m
f
or
d
a
t
a
c
om
m
u
n
ic
at
io
n
In
th
e
da
ta
c
om
m
uni
c
a
ti
on
pha
s
e
,
r
e
c
e
iv
e
d
da
ta
is
tr
a
ns
m
it
te
d
to
an
onboa
r
d
uni
t
by
s
e
le
c
ti
ng
th
e
opt
im
a
l
pa
th
us
in
g
a
c
ogni
ti
ve
le
a
r
ni
ng
-
ba
s
e
d
M
O
C
L
E
A
R
a
lg
or
it
hm
.
E
f
f
e
c
ti
ve
a
nd
lo
s
s
le
s
s
da
ta
c
om
m
uni
c
a
ti
on
is
a
r
c
hi
ve
d
w
it
h
th
e
he
lp
of
th
e
M
O
C
L
E
A
R
a
l
gor
it
hm
,
w
hi
c
h
is
a
c
om
bi
na
ti
on
of
I
M
O
a
nd
opt
im
a
l
c
lu
s
te
r
he
a
d
node
it
e
r
a
ti
ve
ly
in
th
e
I
oT
c
lo
ud
ne
twor
k
m
ode
l.
I
M
O
,
th
e
a
lg
or
it
hm
im
pl
e
m
e
nt
s
gr
e
y
w
ol
f
hunt
in
g
be
ha
vi
our
c
onc
e
pt
s
.
It
ha
s
th
r
e
e
s
ta
ge
s
:
a
ppr
oa
c
h,
hunt
,
a
nd
a
tt
a
c
k
s
ta
g
e
s
.
A
c
ogni
ti
ve
e
xpe
r
t
pos
it
io
n
is
c
r
e
a
te
d,
a
nd
a
f
te
r
pos
it
io
ni
ng,
e
ve
r
y
le
gi
ti
m
a
te
m
ove
ge
ts
upda
te
d.
I
nt
e
ll
ig
e
nt
m
ul
ti
-
obj
e
c
ti
ve
opt
im
iz
a
ti
on
ha
s
f
our
s
e
t
s
ol
ut
io
ns
,
a
lp
h
a
,
be
ta
,
om
e
ga
,
a
nd
de
lt
a
,
w
it
h
th
e
e
xpe
r
t
po
s
it
io
n.
E
xpe
r
t
pos
it
io
ns
a
r
e
a
ls
o
upd
a
te
d
w
it
h
a
gr
a
vi
ta
ti
ona
l
s
e
a
r
c
h
a
lg
or
it
hm
.
T
h
e
M
O
C
L
E
A
R
a
lg
or
it
hm
is
th
e
m
ode
l
th
a
t
de
c
id
e
s
th
e
be
s
t
r
out
e
f
or
da
ta
c
om
m
uni
c
a
ti
on
to
th
e
bo
a
r
d
uni
t.
T
hi
s
hybr
id
m
ode
l
is
th
e
m
os
t
a
dva
nt
a
g
e
ous
f
or
f
in
di
ng
a
num
be
r
of
r
out
e
s
f
or
da
ta
c
om
m
uni
c
a
ti
on
us
in
g
c
on
ve
r
ge
nc
e
.
T
he
upda
ti
ng
pr
oc
e
s
s
a
r
c
hi
ve
d
th
e
w
a
y
th
a
t
te
r
m
s
a
r
e
in
c
lu
de
d
in
th
e
I
M
O
opt
im
iz
a
ti
on
te
c
hni
que
w
it
h
th
e
he
lp
of
th
e
I
oT
c
lo
ud
ne
twor
k
a
lg
or
it
hm
.
M
odi
f
ie
d
M
O
C
L
E
A
R
e
xpr
e
s
s
io
n
a
s
i
n (
2)
.
(
+
1
)
=
1
+
2
+
3
+
4
4
(
2)
W
he
r
e
1
,
2
,
a
nd
G
W
3
in
di
c
a
te
th
e
in
it
ia
l
a
nd
c
ont
in
ue
d
pos
it
io
ns
of
gr
e
y
im
a
ge
s
.
A
ls
o,
4
de
not
e
s
,
at
th
e
ti
m
e
,
gr
e
y
im
a
ge
s
e
nd
pos
it
io
n
w
it
h
M
O
C
L
E
A
R
.
F
ur
th
e
r
gr
e
y
im
a
ge
s
r
e
pr
e
s
e
nt
a
ti
on
a
s
i
n
(
3)
-
(
5)
.
GW
1
=
GW
α
−
I
1
(
D
α
)
(
3)
2
=
−
2
(
)
(
4)
3
=
−
1
(
)
(
5)
W
he
r
e
,
a
nd
de
not
e
s
th
e
opt
im
a
l
s
ol
ut
io
n.
I
ndi
c
a
te
s
th
e
ga
p
be
twe
e
n
th
e
pos
it
io
n
a
nd
th
e
e
va
lu
a
te
d
e
xpe
r
t
pos
it
io
n.
r
e
pr
e
s
e
nt
s
th
e
r
e
ve
r
s
e
of
,
i.
e
.,
th
e
ga
p
be
twe
e
n
th
e
e
va
lu
a
t
e
d
pos
it
io
n
a
nd
th
e
e
xpe
r
t
pos
it
io
n
is
,
w
hi
c
h
in
di
c
a
te
s
th
e
di
s
ta
n
c
e
be
twe
e
n
c
o
gni
ti
ve
e
xpe
r
t
a
nd
th
e
pr
e
di
c
ti
on
-
ba
s
e
d
pos
it
io
n.
4
in
di
c
a
te
s
th
e
pos
it
io
n
of
F
r
a
c
ti
ona
l
G
r
e
y
S
e
a
r
c
h
a
lg
or
i
th
m
w
it
h
a
time
of
t,
e
s
ti
m
a
ti
on
e
qua
ti
on
a
s
i
n (
6)
.
4
=
(
)
+
(
+
1
)
+
1
2
(
−
1
)
(
6)
W
he
r
e
(
)
de
not
e
s
th
e
pos
it
io
n
of
e
xpe
r
t
I
at
th
e
ℎ
f
a
c
to
r
at
a
ti
m
e
,
w
he
r
e
a
s
th
e
e
xpe
r
t’
s
e
v
a
lu
a
te
d
f
r
ont
pos
it
io
n
at
ℎ
ℎ
di
m
e
ns
io
n
in
he
t
−
1
ℎ
time
is
in
di
c
a
te
d
by
(
−
1
)
,
he
r
e
,
(
+
1
)
de
not
e
s
ve
lo
c
it
y
w
it
h
(
+
1
)
ℎ
lo
c
a
ti
on
w
hi
le
r
e
pr
e
s
e
nt
s
pr
oba
bi
li
ty
va
lu
e
s
f
r
om
0
to
1.
T
hus
,
r
ic
e
pl
a
nt
da
ta
ga
in
e
d
th
r
ough
s
e
ns
or
s
or
c
a
m
e
r
a
s
a
nd
ot
he
r
c
om
m
uni
c
a
ti
on
de
vi
c
e
s
a
r
e
gi
v
e
n
to
an
onboa
r
d
uni
t,
w
h
e
r
e
r
ic
e
pl
a
nt
di
s
e
a
s
e
de
te
c
ti
on
a
nd
pr
e
ve
nt
io
n
is
im
pl
e
m
e
nt
e
d.
3.4
.
I
d
e
n
t
if
ic
at
io
n
/
d
e
t
e
c
t
io
n
of
r
ic
e
c
r
op
d
is
e
a
s
e
s
T
he
r
ic
e
c
r
op
di
s
e
a
s
e
de
te
c
ti
on
a
nd
pr
e
ve
nt
io
n
w
it
h
s
e
ns
e
d
r
ic
e
c
r
op
in
f
or
m
a
ti
on
th
r
ough
I
oT
.
A
f
te
r
r
e
c
e
iv
in
g
va
lu
e
s
or
im
a
ge
s
f
ol
lo
w
e
d
by
th
e
pr
oc
e
s
s
e
s
a
r
e
im
a
ge
tr
a
ns
f
or
m
a
ti
on,
de
noi
s
in
g,
d
a
ta
r
e
duc
ti
on,
im
a
ge
s
e
gm
e
nt
a
ti
on,
f
e
a
tu
r
e
s
e
xt
r
a
c
ti
on,
e
nha
n
c
e
m
e
nt
,
i
m
a
ge
c
la
s
s
if
ic
a
ti
on
w
it
h
opt
im
iz
a
ti
on,
a
nd
pe
r
f
or
m
a
nc
e
c
a
lc
ul
a
ti
on
f
or
f
ut
u
r
e
m
a
c
hi
ne
tr
a
in
in
g
pur
pos
e
s
.
I
m
a
ge
s
e
gm
e
nt
a
ti
on
is
a
c
hi
e
ve
d
w
it
h
C
R
C
N
e
t
a
r
c
hi
te
c
tu
r
e
f
or
im
a
ge
lo
c
a
li
z
a
ti
on.
It
is
a
tt
a
in
e
d
w
it
h
s
e
gm
e
nt
s
th
a
t
s
uppor
t
s
ig
ni
f
ic
a
nt
a
nd
e
f
f
ic
ie
nt
obj
e
c
t
de
te
c
ti
on
of
f
e
a
tu
r
e
e
xt
r
a
c
ti
on.
F
in
a
ll
y,
L
I
M
O
pr
opos
e
d
a
f
r
a
m
e
w
or
k
th
a
t
w
a
s
p
e
r
f
or
m
e
d
w
it
h
th
e
G
A
N
ne
twor
k
M
O
C
L
E
A
R
a
nd
tr
a
in
e
d
th
e
m
ode
l
w
it
h
th
e
L
I
M
O
a
lg
or
it
hm
.
A
de
ta
il
e
d
de
s
ig
n
of
L
I
M
O
a
nd
M
O
C
L
E
A
R
f
or
r
ic
e
c
r
op
di
s
e
a
s
e
d
e
te
c
ti
on
w
il
l
be
di
s
c
us
s
e
d
in
th
e
im
pe
ndi
ng
pos
it
io
n
of
th
is
r
e
s
e
a
r
c
h
w
or
k.
C
ons
id
e
r
th
e
in
te
ll
e
c
tu
a
l
s
e
gm
e
nt
s
c
r
e
a
te
d
f
r
om
in
put
c
r
op
im
a
ge
s
,
w
he
r
e
it
de
not
e
s
ove
r
a
ll
s
e
gm
e
nt
s
s
ur
vi
vi
ng
on
in
put
r
ic
e
pl
a
nt
im
a
ge
,
r
e
pr
e
s
e
nt
s
ℎ
ℎ
s
e
gm
e
nt
of
th
e
in
put
.
=
{
1
,
2
,
.
.
.
,
,
.
.
.
,
}
(
7)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
L
aur
e
nt
s
e
r
i
e
s
i
nt
e
ll
ig
e
nt
m
ul
ti
di
m
e
ns
io
nal
obj
e
c
t
opt
imi
z
at
io
n c
la
s
s
if
ic
at
io
n
…
(
A
nandhan
K
ar
unanit
hi
)
4055
3.4.1
.
L
au
r
e
n
t
s
e
r
ie
s
e
xp
r
e
s
s
io
n
T
he
L
a
ur
e
nt
s
e
r
ie
s
w
it
h
c
om
pl
e
x
c
oe
f
f
ic
ie
nt
s
can
be
us
e
d
to
s
tu
dy
th
e
be
ha
vi
or
of
f
unc
ti
ons
ne
a
r
in
di
vi
dua
li
ti
e
s
,
e
s
pe
c
ia
ll
y
in
c
om
pl
e
x
a
na
ly
s
is
.
A
la
ur
e
nt
pol
yn
om
ia
l
is
a
la
ur
e
nt
s
e
r
ie
s
in
w
hi
c
h
onl
y
a
f
in
it
e
num
be
r
of
c
oe
f
f
ic
ie
nt
s
a
r
e
non
-
z
e
r
o.
T
he
L
a
ur
e
nt
pol
yno
m
ia
l
di
f
f
e
r
s
f
r
om
nor
m
a
l
pol
ynom
ia
ls
in
th
a
t
it
can
ha
ve
ne
ga
ti
ve
de
gr
e
e
te
r
m
s
.
W
he
r
e
f
(
x)
is
L
a
ur
e
nt
s
e
r
ie
s
c
om
pl
e
x
f
unc
ti
on,
a
nd
y
is
c
ons
ta
nt
w
it
h
a
n
de
f
in
e
d
by
a
c
ont
our
in
te
gr
a
l.
(
)
=
∑
(
−
)
∞
=
−
∞
(
8)
3.5
.
L
I
M
O
w
it
h
G
A
N
n
e
t
w
or
k
f
or
r
i
c
e
c
r
op
d
is
e
a
s
e
p
r
e
v
e
n
t
io
n
an
d
d
e
t
e
c
t
io
n
T
hi
s
pr
opos
e
d
L
I
M
O
f
r
a
m
e
w
or
k
a
im
s
to
id
e
nt
if
y
th
e
di
s
e
a
s
e
of
r
ic
e
c
r
ops
w
it
h
in
te
ll
e
c
tu
a
l
m
ul
ti
-
obj
e
c
t
opt
im
iz
a
ti
on
a
nd
M
O
C
L
E
A
R
.
L
a
ur
e
nt
s
e
r
ie
s
im
pl
e
m
e
nt
s
th
e
m
e
th
ods
of
c
om
pl
e
x
va
r
ia
bl
e
s
w
it
h
th
e
a
ddi
ti
on
of
in
f
in
it
e
te
r
m
s
a
nd
e
xt
e
ns
io
ns
.
F
or
im
a
ge
pr
oc
e
s
s
in
g
a
nd
obj
e
c
t
de
te
c
ti
on,
th
is
L
I
M
O
f
r
a
m
e
w
or
k
is
an
e
f
f
e
c
ti
ve
a
nd
pow
e
r
f
ul
de
te
c
ti
on
m
e
th
od
be
c
a
us
e
of
c
om
pu
ta
ti
ona
l
in
te
gr
a
ti
on
a
nd
in
f
in
it
e
s
um
c
r
e
a
ti
on.
M
or
e
ove
r
,
th
is
L
I
M
O
s
e
r
ie
s
is
a
one
-
s
te
p
pr
oc
e
s
s
of
m
oni
to
r
in
g
w
it
h
hi
gh
-
di
m
e
ns
io
na
l
te
r
m
s
.
T
hi
s
L
a
ur
e
nt
s
e
r
ie
s
s
uppor
ts
th
e
de
r
iv
in
g
of
im
a
gi
na
r
y
uppe
r
e
r
r
or
a
nd
c
on
ve
r
ge
nc
e
.
T
he
I
M
O
m
ode
l
m
a
in
ta
in
s
s
ta
bi
li
ty
be
twe
e
n
m
a
ni
pul
a
ti
on
a
nd
in
ve
s
ti
ga
ti
on.
I
M
O
,
L
a
ur
e
nt
s
e
r
io
us
ly
in
c
r
e
a
s
e
s
th
e
opt
im
a
l
w
a
y
f
or
c
onj
un
c
ti
on
pr
oc
e
s
s
e
s
,
th
e
m
os
t
f
a
vor
a
bl
e
gr
ound
s
ol
ut
io
ns
,
a
nd
th
e
s
y
m
m
e
tr
y
be
twe
e
n
tr
a
in
e
d
a
nd
te
s
t
da
ta
.
H
e
r
e
,
in
te
gr
a
te
d
I
M
O
opt
im
iz
a
ti
on
a
nd
th
e
L
a
ur
e
nt
s
e
r
ie
s
a
r
e
us
e
d
to
r
a
is
e
th
e
ove
r
a
ll
th
r
oughput
s
ig
ni
f
ic
a
nt
ly
.
F
ig
ur
e
2
s
how
s
th
e
f
ol
lo
w
in
g
G
A
N
w
it
h
L
I
M
O
ne
twor
k
a
nd
its
a
r
c
hi
te
c
tu
r
e
us
e
d
in
th
is
r
ic
e
pl
a
nt
di
s
e
a
s
e
de
te
c
ti
on
s
y
s
te
m
.
G
A
N
ne
twor
ks
in
vol
ve
two
ty
pe
s
of
n
e
ur
a
l
n
e
twor
k
m
ode
ls
:
ge
n
e
r
a
to
r
s
a
nd
di
s
c
r
im
in
a
to
r
s
.
G
e
ne
r
a
to
r
s
m
a
ke
pr
e
di
c
ti
ons
or
a
ppr
oxi
m
a
te
s
a
m
pl
e
s
on
an
or
ig
in
a
l
or
e
xpe
c
te
d
out
c
om
e
ba
s
is
.
D
is
c
r
im
in
a
to
r
f
in
ds
th
e
va
r
ia
ti
on
f
r
om
r
e
gul
a
r
a
c
ti
vi
ti
e
s
.
T
hi
s
pr
oc
e
s
s
be
twe
e
n
ge
ne
r
a
to
r
a
nd
di
s
c
r
im
in
a
to
r
m
ode
ls
le
a
ds
up
to
th
e
le
v
e
ls
of
pe
r
f
e
c
ti
on.
V
a
r
io
us
opt
im
um
va
lu
e
s
a
r
e
r
e
c
e
iv
e
d
th
r
ough
di
f
f
e
r
e
nt
a
lg
or
it
hm
s
th
a
t
s
uppor
t
th
e
tr
a
in
in
g
of
c
r
op
di
s
e
a
s
e
im
a
ge
s
a
nd
th
e
c
om
pi
la
ti
on
of
th
e
f
r
a
m
e
w
or
k.
F
ig
ur
e
2.
R
ic
e
pl
a
nt
d
is
e
a
s
e
de
te
c
ti
on
u
s
in
g
L
I
M
O
w
it
h
G
A
N
f
r
a
m
e
w
or
k
4.
R
E
S
U
L
T
S
AND
D
I
S
C
U
S
S
I
O
N
O
ur
r
e
s
e
a
r
c
h
w
or
k
in
ve
s
ti
ga
te
d
th
e
e
f
f
e
c
ts
of
L
I
M
O
c
la
s
s
if
ic
a
ti
on
a
nd
M
O
C
L
E
A
R
s
e
gm
e
nt
a
ti
on
on
r
ic
e
c
r
op
di
s
e
a
s
e
de
t
e
c
ti
on.
W
hi
le
e
a
r
li
e
r
s
tu
di
e
s
h
a
ve
e
xpl
or
e
d
th
e
im
pa
c
t
of
de
e
p
le
a
r
ni
ng
-
ba
s
e
d
c
la
s
s
if
ic
a
ti
on
m
e
th
ods
a
nd
I
oT
-
ba
s
e
d
di
s
e
a
s
e
m
oni
to
r
in
g,
th
e
y
ha
ve
not
e
xpl
ic
it
ly
a
ddr
e
s
s
e
d
it
s
in
f
lu
e
nc
e
on
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 5, O
c
to
be
r
2025
:
4050
-
4060
4056
th
e
in
te
gr
a
ti
on
of
th
e
L
a
ur
e
nt
s
e
r
ie
s
w
it
h
I
M
O
f
or
e
nha
nc
e
d
d
is
e
a
s
e
de
te
c
ti
on
a
c
c
ur
a
c
y.
We
f
ound
th
a
t
th
e
in
te
gr
a
ti
on
of
L
I
M
O
c
la
s
s
if
ic
a
ti
on
w
it
h
M
O
C
L
E
A
R
s
e
gm
e
nt
a
t
io
n
c
or
r
e
la
te
s
w
it
h
e
nha
nc
e
d
a
c
c
ur
a
c
y
in
r
ic
e
c
r
op
di
s
e
a
s
e
de
t
e
c
ti
on.
O
ur
pr
opos
e
d
m
e
th
od
in
th
is
s
tu
dy
pr
ovi
de
s
a
tr
e
m
e
ndous
ly
a
dva
nc
e
d
pr
opor
ti
on
of
c
or
r
e
c
tl
y
c
la
s
s
if
ie
d
di
s
e
a
s
e
in
s
ta
nc
e
s
c
om
pa
r
e
d
to
m
is
c
la
s
s
if
ie
d
c
a
s
e
s
.
F
ig
ur
e
3
s
how
s
e
xpe
r
im
e
nt
a
l
a
na
ly
s
is
us
in
g
L
I
M
O
w
it
h
G
A
N
ne
twor
k,
F
ig
u
r
e
3(
a
)
e
xpl
a
in
s
th
e
va
r
io
us
r
ic
e
c
r
op
di
s
e
a
s
e
im
a
ge
s
obt
a
in
e
d
f
r
om
th
e
r
ic
e
c
r
op
di
s
e
a
s
e
da
t
a
s
e
t.
F
ig
ur
e
3(
b)
s
how
s
an
im
a
ge
pr
e
-
pr
oc
e
s
s
in
g
w
it
h
da
ta
r
e
duc
ti
on
a
nd
f
e
a
tu
r
e
e
ngi
ne
e
r
in
g
c
ont
a
in
e
d
f
r
om
in
put
va
lu
e
s
.
F
ig
ur
e
3(
c
)
de
s
c
r
ib
e
s
th
e
gr
ound
tr
ut
h
va
lu
e
s
of
in
put
im
a
ge
s
s
e
gm
e
nt
e
d
in
put
obt
a
in
e
d
bounda
r
y
va
lu
e
s
by
C
R
C
N
e
t
is
r
e
pr
e
s
e
nt
e
d
in
F
ig
ur
e
3(
d)
f
in
a
l
s
e
gm
e
nt
a
ti
on
pr
oc
e
s
s
.
T
hi
s
pe
r
f
or
m
a
nc
e
a
na
ly
s
is
f
oc
us
e
d
on
c
a
lc
ul
a
ti
ng
hi
ghl
y
a
c
c
ur
a
te
pr
e
di
c
ti
on
a
nd
r
e
c
a
ll
w
it
h
th
e
pr
opos
e
d
L
I
M
O
in
th
e
G
A
N
ne
twor
k
to
e
xpr
e
s
s
its
e
f
f
ic
a
c
y
in
s
ta
ndi
ngs
of
a
c
c
ur
a
c
y,
r
e
c
a
ll
,
pr
e
c
is
io
n,
s
e
ns
it
iv
it
y,
a
nd
s
pe
c
if
ic
it
y
m
e
tr
ic
s
.
T
hi
s
pe
r
f
or
m
a
nc
e
a
na
ly
s
is
pr
oc
e
s
s
ha
s
di
ve
r
s
e
di
m
e
ns
io
na
li
ti
e
s
,
s
uc
h
as
va
r
io
us
e
poc
h
va
lu
e
s
,
s
u
c
h
as
2
,
500
to
3
,
400,
a
nd
di
f
f
e
r
e
nt
ba
tc
h
s
iz
e
s
(
r
a
ngi
ng
f
r
om
80
to
220)
.
T
h
e
pr
opos
e
d
m
e
th
od
c
oul
d
pot
e
nt
ia
ll
y
e
nha
nc
e
th
e
a
c
c
ur
a
c
y
of
c
r
op
di
s
e
a
s
e
c
la
s
s
if
ic
a
ti
on
w
it
hout
ne
g
a
ti
ve
ly
a
f
f
e
c
ti
ng
th
e
m
ode
l'
s
a
bi
li
ty
to
d
is
ti
ngui
s
h
be
twe
e
n
di
f
f
e
r
e
nt
di
s
e
a
s
e
ty
pe
s
.
F
ig
ur
e
4
s
how
s
th
e
e
xpe
r
im
e
nt
a
l
r
ic
e
pl
a
nt
di
s
e
a
s
e
r
e
s
ul
ts
c
onqu
e
r
e
d
w
it
h
th
e
L
I
M
O
f
r
a
m
e
w
or
k
(
L
a
ur
e
nt
s
e
r
ie
s
-
I
M
O
)
ba
s
e
d
G
A
N
ne
twor
k
us
in
g
th
e
r
ic
e
c
r
op
di
s
e
a
s
e
da
ta
s
e
t.
T
h
e
e
xpe
r
im
e
nt
a
l
a
na
ly
s
is
of
L
I
M
O
w
it
h
th
e
ge
ne
r
a
ti
ve
ne
twor
k
to
c
a
lc
ul
a
te
th
e
p
e
r
c
e
nt
a
ge
a
c
c
ur
a
c
y
r
a
te
,
s
e
ns
it
iv
it
y
va
lu
e
s
,
a
nd
s
pe
c
if
ic
it
y
va
lu
e
s
us
in
g
va
r
io
us
e
poc
h
v
a
lu
e
s
.
T
hi
s
s
tu
dy
e
xa
m
in
e
d
a
c
om
pr
e
he
ns
iv
e
L
I
M
O
-
ba
s
e
d
c
la
s
s
if
ic
a
t
io
n
f
r
a
m
e
w
or
k
f
or
r
ic
e
c
r
op
di
s
e
a
s
e
d
e
te
c
ti
on
w
it
h
G
A
N
-
e
nha
nc
e
d
s
e
gm
e
nt
a
ti
on
te
c
hni
que
s
.
H
ow
e
ve
r
,
m
or
e
th
or
ough
r
e
s
e
a
r
c
h
m
a
y
be
r
e
qui
r
e
d
to
va
li
da
te
its
ge
ne
r
a
li
z
a
bi
li
ty
a
c
r
os
s
di
f
f
e
r
e
nt
c
r
op
ty
pe
s
,
pa
r
ti
c
ul
a
r
ly
in
r
e
la
ti
on
to
va
r
ia
ti
ons
in
e
nvi
r
onm
e
nt
a
l
c
ondi
ti
ons
a
nd
im
a
ge
qua
li
ty
.
(
a
)
(
b)
(
c
)
(
d)
F
ig
ur
e
3.
E
xpe
r
im
e
nt
a
l
a
na
ly
s
is
of
th
e
L
I
M
O
c
om
bi
ne
d
w
it
h
GAN
ne
twor
k
:
(
a)
r
e
a
l
in
put
im
a
ge
,
(
b)
m
ul
ti
-
ob
je
c
ti
ve
ba
s
e
d
pr
e
-
pr
oc
e
s
s
in
g
in
put
im
a
ge
,
(
c)
gr
ound
tr
ut
h
in
put
im
a
ge
, a
nd
(
d)
L
I
M
O
s
e
gm
e
nt
a
ti
on
por
ti
on
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
L
aur
e
nt
s
e
r
i
e
s
i
nt
e
ll
ig
e
nt
m
ul
ti
di
m
e
ns
io
nal
obj
e
c
t
opt
imi
z
at
io
n c
la
s
s
if
ic
at
io
n
…
(
A
nandhan
K
ar
unanit
hi
)
4057
In
th
is
e
xpe
r
im
e
nt
a
l
a
na
ly
s
is
,
F
ig
ur
e
4(
a
)
s
how
s
th
e
s
pe
c
if
ic
it
y
f
a
c
to
r
s
e
va
lu
a
ti
on
of
th
e
p
r
opos
e
d
c
la
s
s
if
ic
a
ti
on
m
ode
l
w
it
h
ot
he
r
e
xi
s
ti
ng
m
ode
ls
.
At
th
e
time
8
5%
of
th
e
tr
a
in
in
g
p
r
oc
e
s
s
w
a
s
c
om
pl
e
te
d,
th
e
s
pe
c
if
ic
it
y
pe
r
c
e
nt
a
g
e
of
th
e
L
I
M
O
m
ode
l
w
a
s
a
tt
a
in
e
d
at
onl
y
92.4%
.
a
ls
o,
ot
he
r
m
ode
ls
a
r
e
D
e
e
p
N
e
ur
a
l
N
e
twor
k
(
DNN
)
w
a
s
80.4,
C
N
N
w
a
s
85.3,
Y
O
L
O
v
2
w
a
s
84.3,
a
nd
R
e
s
N
e
t
w
a
s
88.6%
[
23]
–
[
25]
.
F
ig
ur
e
4
(
b)
e
xpl
a
in
s
th
e
c
om
pa
r
is
on
of
th
e
s
e
n
s
it
iv
it
y
f
a
c
to
r
s
of
L
I
M
O
w
it
h
ot
he
r
e
xi
s
ti
ng
m
ode
ls
.
At
th
e
time
of
th
e
85%
tr
a
in
in
g
s
e
t
c
om
pl
e
te
d,
th
e
s
e
ns
it
iv
it
y
r
a
te
w
a
s
L
I
M
O
=
92.6,
D
N
N
=
73.1,
C
N
N
=
90.2,
Y
O
L
O
v
2=
68.8,
a
nd
R
e
s
N
e
t=
83.4%
.
F
ig
ur
e
4(
c
)
por
tr
a
ys
50
to
85%
of
tr
a
in
in
g
da
ta
a
tt
a
in
e
d
a
c
c
ur
a
c
y
le
v
e
ls
w
it
h
th
e
L
I
M
O
f
r
a
m
e
w
or
k
a
nd
ot
he
r
e
xi
s
ti
ng
m
ode
ls
.
At
th
e
e
nd
of
th
e
a
na
l
ys
is
,
w
he
n
th
e
85%
tr
a
in
in
g
s
e
t
is
c
om
pl
e
te
d,
th
e
a
c
c
ur
a
c
y
pe
r
c
e
nt
a
g
e
is
L
I
M
O
w
it
h
G
A
N
=
91.5,
D
N
N
=
70.2,
C
N
N
=
80.4,
Y
O
L
O
v
2=
69.7,
a
nd
R
e
s
N
e
t1
50
=
80.3%
.
T
hr
ough
th
is
pr
opos
e
d
r
e
s
e
a
r
c
h
w
or
k,
L
I
M
O
s
uppor
ts
id
e
nt
if
yi
ng
c
r
op
di
s
e
a
s
e
s
a
nd
th
e
ir
le
ve
ls
ve
r
y
e
f
f
e
c
ti
ve
ly
,
w
hi
c
h
he
lp
s
to
m
a
ke
a
dva
nc
e
d
s
m
a
r
t
a
g
r
ic
ul
tu
r
e
f
or
be
tt
e
r
pr
oduc
ti
vi
ty
.
A
c
c
or
di
ng
to
our
s
tu
dy,
a
hi
ghe
r
s
e
n
s
it
iv
it
y
in
di
s
e
a
s
e
d
e
te
c
ti
on
doe
s
not
n
e
c
e
s
s
a
r
il
y
in
di
c
a
t
e
poor
s
p
e
c
if
ic
it
y
pe
r
f
or
m
a
nc
e
.
T
he
pr
opos
e
d
m
e
th
od
c
oul
d
pot
e
nt
ia
ll
y
e
nha
nc
e
th
e
a
c
c
ur
a
c
y
of
c
r
op
di
s
e
a
s
e
c
la
s
s
if
ic
a
ti
on
w
it
hout
ne
ga
ti
ve
ly
a
f
f
e
c
ti
ng
th
e
m
ode
l'
s
a
bi
li
ty
to
di
s
ti
ngui
s
h
be
twe
e
n
d
if
f
e
r
e
nt
di
s
e
a
s
e
ty
pe
s
.
(
a
)
(
b)
(
c
)
F
ig
ur
e
4.
E
xpe
r
im
e
nt
a
l
a
na
ly
s
is
of
50%
to
85%
tr
a
in
in
g
s
e
t
of
d
e
te
c
ti
on
(
a)
s
pe
c
if
ic
it
y,
(
b)
s
e
ns
it
iv
it
y,
a
nd
(
c)
a
c
c
ur
a
c
y
by
L
I
M
O
w
it
h
G
A
N
m
ode
l
c
om
pa
r
e
d
w
it
h
va
r
io
us
e
xi
s
ti
ng
m
ode
ls
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 5, O
c
to
be
r
2025
:
4050
-
4060
4058
T
a
bl
e
2
da
ta
pr
ovi
de
our
pr
opos
e
d
L
I
M
O
m
ode
l,
w
it
h
th
e
G
A
N
ne
twor
k
ge
tt
in
g
th
e
hi
ghe
s
t
a
c
c
ur
a
c
y
r
a
te
is
91.5
%
,
w
he
r
e
a
s
th
e
a
c
c
ur
a
c
y
r
a
te
c
om
pa
r
e
d
w
it
h
e
xi
s
ti
ng
c
la
s
s
if
ic
a
ti
on
m
ode
ls
obt
a
in
e
d
by
DNN,
C
N
N
,
Y
O
L
O
v
2,
a
nd
R
e
s
N
e
t
w
e
r
e
70.2
%
,
80.4
%
,
69.7
%
,
a
nd
80.3
%
[
19]
,
[
26]
,
[
27]
.
A
c
om
pa
r
is
on
of
th
e
s
e
a
c
c
ur
a
c
y
va
lu
e
s
r
e
ve
a
ls
our
pr
opos
e
d
L
I
M
O
f
r
a
m
e
w
or
k
pr
ovi
de
s
34.7%
be
tt
e
r
th
a
n
th
e
de
e
p
ne
ur
a
l
ne
twor
ks
m
ode
l.
E
xpe
r
im
e
nt
a
l
r
e
s
ul
ts
s
how
th
a
t
L
I
M
O
w
it
h
a
G
A
N
ne
twor
k
can
e
nl
a
r
ge
th
e
c
la
s
s
if
ic
a
ti
on
a
c
c
ur
a
c
y
r
a
te
a
nd
c
ont
r
ol
ove
r
f
it
ti
ng
is
s
ue
s
.
T
he
a
c
c
ur
a
c
y
a
n
d
pr
e
di
c
ti
on
r
a
te
ha
ve
s
ig
ni
f
ic
a
nt
ly
im
pr
ove
d
s
in
c
e
our
r
e
s
e
a
r
c
h
w
or
k
pr
ovi
de
s
to
ta
ll
y
di
f
f
e
r
e
nt
s
ugge
s
ti
ons
f
r
om
e
xi
s
ti
ng
G
A
N
-
onl
y
m
ode
ls
.
T
he
s
e
G
A
N
ne
twor
ks
a
r
e
us
e
d
to
im
pr
ove
im
a
ge
r
e
s
ol
ut
i
on
va
lu
e
s
c
ont
r
ol
le
d
w
it
h
c
ogni
ti
ve
le
a
r
ni
ng
m
e
th
ods
.
T
a
bl
e
2
c
om
pa
r
e
s
th
e
act
of
our
pr
opos
e
d
L
I
M
O
f
r
a
m
e
w
or
k
w
it
h
e
xi
s
ti
ng
m
ode
ls
s
u
c
h
as
DNN,
C
N
N
,
Y
O
L
O
v2,
a
nd
R
e
s
N
e
t
in
te
r
m
s
of
a
c
c
ur
a
c
y,
s
e
ns
it
iv
it
y,
a
nd
s
pe
c
if
ic
it
y
[
28]
.
O
ur
s
tu
dy
de
m
ons
tr
a
te
s
th
a
t
th
e
L
I
M
O
c
la
s
s
if
ic
a
ti
on
f
r
a
m
e
w
or
k
c
om
bi
ne
d
w
it
h
G
A
N
-
ba
s
e
d
s
e
gm
e
nt
a
ti
on
is
m
or
e
r
e
s
il
ie
nt
th
a
n
tr
a
di
ti
ona
l
de
e
p
le
a
r
ni
ng
m
ode
ls
in
de
te
c
ti
ng
r
ic
e
c
r
op
di
s
e
a
s
e
s
[
29]
,
[
30]
.
O
ur
s
tu
dy
s
ugge
s
ts
th
a
t
th
e
pr
opos
e
d
f
r
a
m
e
w
or
k
can
tr
a
ns
f
or
m
a
gr
ic
ul
tu
r
a
l
di
s
e
a
s
e
m
a
na
g
e
m
e
nt
by
e
nh
a
nc
in
g
de
t
e
c
ti
on
a
c
c
ur
a
c
y,
e
ns
ur
in
g
r
e
a
l
-
time
m
oni
to
r
in
g,
a
nd
e
n
a
bl
in
g
m
or
e
e
f
f
ic
ie
nt
de
c
is
io
n
-
m
a
ki
ng.
T
he
s
e
a
dva
n
c
e
m
e
nt
s
not
onl
y
im
pr
ove
c
r
op
y
ie
ld
a
nd
qua
li
ty
but
a
ls
o
c
ont
r
ib
ut
e
to
t
he
br
oa
de
r
goa
l
of
s
us
ta
in
a
bl
e
a
nd
te
c
hnol
ogy
-
dr
iv
e
n
a
gr
ic
ul
tu
r
e
a
nd
r
e
a
l
-
time
di
s
e
a
s
e
pr
e
di
c
ti
on
a
nd
m
oni
to
r
in
g.
T
a
bl
e
2.
O
ve
r
a
ll
pe
r
f
or
m
a
nc
e
a
na
ly
s
is
of
a
c
c
ur
a
c
y,
s
e
n
s
it
iv
it
y,
a
nd
s
pe
c
if
ic
it
y
P
e
r
f
or
m
a
nc
e
a
na
l
ys
i
s
pe
r
c
e
nt
a
ge
(%)
DNN
C
N
N
Y
O
L
O
v
2
R
e
s
N
e
t
L
I
M
O
w
i
t
h
GAN
D
a
t
a
s
e
t
c
ount
5
,
000
3
,
500
2
,
650
1
,
550
2
,
400
A
c
c
ur
a
c
y
70.2
80.4
69.7
80.3
91.5
S
e
ns
i
t
i
vi
t
y
73.1
90.2
68.8
83.4
92.6
S
pe
c
i
f
i
c
i
t
y
80.4
85.3
84.3
88.6
92.4
5.
C
O
N
C
L
U
S
I
O
N
L
a
ur
e
nt
s
e
r
ie
s
-
I
M
O
w
it
h
G
A
N
ne
twor
k
f
or
r
ic
e
c
r
op
di
s
e
a
s
e
id
e
nt
if
ic
a
ti
on
a
nd
pr
e
ve
nt
io
n
in
th
e
I
oT
pl
a
tf
or
m
.
T
hr
ough
th
is
f
r
a
m
e
w
or
k,
th
e
I
oT
ne
twor
k
node
s
c
ol
le
c
t
va
lu
e
s
f
r
om
th
e
r
ic
e
c
r
op
le
a
ve
s
th
r
ough
th
e
s
e
ns
or
,
a
lt
hough
th
e
s
e
ns
e
d
va
lu
e
s
a
r
e
c
om
m
uni
c
a
te
d
w
it
h
th
e
ba
s
e
s
ta
ti
on
th
r
ough
M
O
C
L
E
A
R
.
T
hi
s
M
O
C
L
E
A
R
m
e
th
od
s
e
le
c
t
e
d
th
e
opt
im
a
l
r
out
e
b
e
twe
e
n
tr
a
ns
m
is
s
io
n
nod
e
s
,
w
hi
c
h
nod
e
s
pa
r
ti
c
ip
a
te
d
in
c
om
m
uni
c
a
ti
on.
T
he
r
e
s
e
a
r
c
h
f
r
a
m
e
w
or
k
w
a
s
m
odi
f
ie
d
on
pr
e
-
pr
oc
e
s
s
in
g
w
it
h
d
a
ta
r
e
duc
ti
on
a
nd
f
e
a
tu
r
e
e
ngi
ne
e
r
in
g
to
im
pr
ove
th
e
qua
li
ty
of
obj
e
c
t
de
te
c
ti
on.
T
h
e
n,
th
e
c
ogni
ti
ve
r
e
s
id
ua
l
c
onvolut
io
na
l
ne
twor
k
w
a
s
r
e
vi
s
e
d
to
s
e
gm
e
nt
th
e
s
e
n
s
e
d
va
lu
e
s
w
it
h
a
ppr
opr
ia
te
f
e
a
tu
r
e
s
e
le
c
ti
ons
to
im
pr
ove
th
e
e
f
f
ic
ie
nc
y
of
c
r
op
di
s
e
a
s
e
id
e
nt
if
ic
a
ti
on.
L
I
M
O
c
la
s
s
if
ic
a
ti
on
w
a
s
im
pl
e
m
e
nt
e
d
us
in
g
th
e
L
a
ur
e
nt
s
e
r
ie
s
a
nd
in
te
ll
ig
e
nt
m
ul
ti
di
m
e
ns
io
na
l
obj
e
c
t
-
ba
s
e
d
opt
im
iz
a
ti
on.
P
e
r
f
or
m
a
nc
e
a
n
a
ly
s
is
of
our
L
I
M
O
m
ode
l
a
c
hi
e
ve
d
be
tt
e
r
r
e
s
ul
ts
s
uc
h
as
a
c
c
ur
a
c
y:
91.5%
,
s
pe
c
if
ic
it
y:
92.41%
,
a
nd
s
e
ns
i
ti
vi
ty
:
92.6%
.
P
a
ddy
le
a
f
im
a
ge
c
la
s
s
if
ic
a
ti
on
w
il
l
be
e
nha
nc
e
d
by
c
ont
in
ue
d
e
xpl
or
a
ti
on
in
th
e
s
e
di
r
e
c
ti
ons
,
w
hi
c
h
w
il
l
a
ddr
e
s
s
r
e
a
l
-
w
or
ld
c
ha
ll
e
nge
s
in
a
gr
ic
ul
tu
r
e
.
F
ut
ur
e
s
tu
di
e
s
in
ve
s
ti
ga
te
th
e
a
da
pt
a
bi
li
ty
of
th
i
s
a
ppr
oa
c
h
to
ot
he
r
c
r
ops
s
pe
c
ie
s
a
nd
e
xpl
or
e
f
e
a
s
ib
le
m
e
th
ods
f
or
pr
oduc
in
g
r
e
a
l
-
time
di
s
e
a
s
e
d
e
te
c
ti
on
s
y
s
t
e
m
s
w
it
h
m
in
im
a
l
c
om
put
a
ti
ona
l
ove
r
he
a
d.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
T
he
a
ut
hor
s
s
t
a
te
no
f
undi
ng
w
a
s
in
vol
ve
d.
A
U
T
H
O
R
C
O
N
T
R
I
B
U
T
I
O
N
S
S
T
A
T
E
M
E
N
T
T
hi
s
jo
ur
na
l
us
e
s
th
e
C
ont
r
ib
ut
or
R
ol
e
s
T
a
xonomy
(
C
R
e
d
iT
)
to
r
e
c
ogni
z
e
in
di
vi
dua
l
a
ut
hor
c
ont
r
ib
ut
io
ns
,
r
e
duc
e
a
ut
hor
s
hi
p
di
s
put
e
s
,
a
nd
f
a
c
il
it
a
te
c
ol
la
bo
r
a
ti
on.
N
am
e
o
f
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
A
na
ndha
n
K
a
r
una
ni
th
i
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
A
ja
y
S
ha
nke
r
S
in
gh
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
l
i
z
a
t
i
on
M
:
M
e
t
hodol
ogy
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
da
t
i
on
Fo
:
Fo
r
m
a
l
a
na
l
ys
i
s
I
:
I
nve
s
t
i
ga
t
i
on
R
:
R
e
s
our
c
e
s
D
:
D
a
t
a
C
ur
a
t
i
on
O
:
W
r
i
t
i
ng
-
O
r
i
gi
na
l
D
r
a
f
t
E
:
W
r
i
t
i
ng
-
R
e
vi
e
w
&
E
di
t
i
ng
Vi
:
Vi
s
ua
l
i
z
a
t
i
on
Su
:
Su
pe
r
vi
s
i
on
P
:
P
r
oj
e
c
t
a
dm
i
ni
s
t
r
a
t
i
on
Fu
:
Fu
ndi
ng a
c
qui
s
i
t
i
on
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
L
aur
e
nt
s
e
r
i
e
s
i
nt
e
ll
ig
e
nt
m
ul
ti
di
m
e
ns
io
nal
obj
e
c
t
opt
imi
z
at
io
n c
la
s
s
if
ic
at
io
n
…
(
A
nandhan
K
ar
unanit
hi
)
4059
C
O
N
F
L
I
C
T
OF
I
N
T
E
R
E
S
T
S
T
A
T
E
M
E
N
T
A
ut
hor
s
s
ta
te
no
c
onf
li
c
t
of
in
te
r
e
s
t.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
T
he
da
ta
th
a
t
s
uppor
t
th
e
f
in
di
ngs
of
th
is
s
tu
dy
a
r
e
a
va
il
a
bl
e
f
r
om
th
e
c
or
r
e
s
ponding
a
ut
hor
,
[
AK
]
,
upon
r
e
a
s
ona
bl
e
r
e
que
s
t.
R
E
F
E
R
E
N
C
E
S
[
1]
A
. F
. J
i
m
e
ne
z
, P
. F
.
C
a
r
de
na
s
,
F
. J
i
m
e
ne
z
,
A
. C
a
n
a
l
e
s
,
a
nd A
.
L
óp
e
z
, “
A
c
ybe
r
-
phys
i
c
a
l
i
nt
e
l
l
i
ge
nt
a
ge
nt
f
or
i
r
r
i
ga
t
i
on s
c
he
dul
i
ng
i
n hor
t
i
c
ul
t
ur
a
l
c
r
ops
,”
C
om
put
e
r
s
and E
l
e
c
t
r
oni
c
s
i
n A
gr
i
c
ul
t
ur
e
, vol
. 178, 202
0, doi
:
10.1016/
j
.c
om
pa
g.2020.105777.
[
2]
A
.
S
a
i
ni
,
N
.
S
.
G
i
l
l
,
P
.
G
ul
i
a
,
A
.
K
.
T
i
w
a
r
i
,
P
.
M
a
r
a
t
ha
,
a
nd
M
.
A
.
S
ha
h,
“
S
m
a
r
t
c
r
op
di
s
e
a
s
e
m
oni
t
or
i
ng
s
ys
t
e
m
i
n
I
oT
us
i
n
g
opt
i
m
i
z
a
t
i
on e
na
bl
e
d de
e
p r
e
s
i
dua
l
ne
t
w
or
k,”
Sc
i
e
nt
i
f
i
c
R
e
por
t
s
, vol
. 15, no. 1,
2025, doi
:
10.1038/
s
41598
-
025
-
85486
-
1.
[
3]
A
.
F
.
J
i
m
e
ne
z
,
P
.
F
.
C
a
r
de
na
s
,
A
.
C
a
na
l
e
s
,
F
.
J
i
m
e
ne
z
,
a
nd
A
.
P
or
t
a
c
i
o,
“
A
s
ur
ve
y
on
i
nt
e
l
l
i
ge
nt
a
ge
nt
s
a
nd
m
ul
t
i
-
a
ge
nt
s
f
or
i
r
r
i
ga
t
i
on s
c
he
dul
i
ng,”
C
om
put
e
r
s
and E
l
e
c
t
r
oni
c
s
i
n A
gr
i
c
ul
t
ur
e
, vol
. 176, 2020, doi
:
10.1016/
j
.c
om
pa
g.2020.105474.
[
4]
Y
.
Z
ha
o
e
t
al
.
,
“
A
n
e
f
f
e
c
t
i
ve
a
ut
om
a
t
i
c
s
y
s
t
e
m
de
pl
oye
d
i
n
a
gr
i
c
ul
t
ur
a
l
i
nt
e
r
ne
t
of
t
hi
ngs
u
s
i
ng
m
ul
t
i
-
c
ont
e
xt
f
us
i
on
n
e
t
w
or
k
t
ow
a
r
ds
c
r
op di
s
e
a
s
e
r
e
c
ogni
t
i
on i
n t
he
w
i
l
d,”
A
ppl
i
e
d Sof
t
C
om
put
i
ng J
our
nal
, vol
. 89, 2020, doi
:
10.1016/
j
.a
s
oc
.2020.106128.
[
5]
R
.
S
ha
r
m
a
,
S
.
S
.
K
a
m
bl
e
,
A
.
G
una
s
e
ka
r
a
n,
V
.
K
um
a
r
,
a
nd
A
.
K
um
a
r
,
“
A
s
ys
t
e
m
a
t
i
c
l
i
t
e
r
a
t
ur
e
r
e
vi
e
w
on
m
a
c
hi
ne
l
e
a
r
ni
ng
a
ppl
i
c
a
t
i
ons
f
or
s
us
t
a
i
na
bl
e
a
gr
i
c
ul
t
ur
e
s
uppl
y
c
ha
i
n
pe
r
f
or
m
a
nc
e
,”
C
om
p
ut
e
r
s
and
O
pe
r
at
i
on
s
R
e
s
e
a
r
c
h
,
vol
.
119,
2020,
doi
:
10.1016/
j
.c
or
.2020.104926.
[
6]
C
.
Z
hou,
Y
.
Z
hong,
S
.
Z
hou,
J
.
S
ong,
a
nd
W
.
X
i
a
ng,
“
R
i
c
e
l
e
a
f
di
s
e
a
s
e
i
de
nt
i
f
i
c
a
t
i
on
by
r
e
s
i
dua
l
-
di
s
t
i
l
l
e
d
t
r
a
ns
f
or
m
e
r
,”
E
ngi
ne
e
r
i
ng A
ppl
i
c
at
i
ons
of
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
, vol
. 121, 2023, doi
:
10.1016
/
j
.e
nga
ppa
i
.2023.106020.
[
7]
A
.
S
ube
e
s
h
a
nd
C
.
R
.
M
e
ht
a
,
“
A
ut
om
a
t
i
on
a
nd
di
gi
t
i
z
a
t
i
on
of
a
gr
i
c
ul
t
ur
e
u
s
i
ng
a
r
t
i
f
i
c
i
a
l
i
nt
e
l
l
i
ge
nc
e
a
nd
i
nt
e
r
ne
t
of
t
hi
ngs
,
”
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
i
n A
gr
i
c
ul
t
ur
e
, vol
. 5, pp. 278
–
291, 2021, doi
:
10.1016/
j
.
a
i
i
a
.2021.11.004.
[
8]
A
.
K
ha
t
t
a
b,
S
.
E
.
D
.
H
a
bi
b,
H
.
I
s
m
a
i
l
,
S
.
Z
a
ya
n,
Y
.
F
a
hm
y,
a
nd
M
.
M
.
K
ha
i
r
y,
“
A
n
I
oT
-
ba
s
e
d
c
ogni
t
i
ve
m
oni
t
or
i
ng
s
ys
t
e
m
f
or
e
a
r
l
y pl
a
nt
di
s
e
a
s
e
f
or
e
c
a
s
t
,”
C
om
put
e
r
s
and E
l
e
c
t
r
oni
c
s
i
n A
gr
i
c
ul
t
ur
e
, vol
. 16
6, 2019, doi
:
10.1016/
j
.c
om
pa
g.2019.105028.
[
9]
M
.
M
.
K
h
a
l
i
d
a
nd
O
.
K
a
r
a
n,
“
D
e
e
p
l
e
a
r
ni
ng
f
or
pl
a
nt
di
s
e
a
s
e
d
e
t
e
c
t
i
on,”
I
nt
e
r
nat
i
onal
J
our
nal
of
M
at
he
m
at
i
c
s
,
St
at
i
s
t
i
c
s
,
and
C
om
put
e
r
Sc
i
e
n
c
e
, vol
. 2, pp. 75
–
84, 2023, doi
:
10.59543/
i
j
m
s
c
s
.v2i
.8343.
[
10]
L
.
Z
hong,
L
.
H
u,
a
nd
H
.
Z
hou,
“
D
e
e
p
l
e
a
r
ni
ng
ba
s
e
d
m
ul
t
i
-
t
e
m
po
r
a
l
c
r
op
c
l
a
s
s
i
f
i
c
a
t
i
on,”
R
e
m
ot
e
Se
ns
i
ng
of
E
nv
i
r
onm
e
nt
,
vol
. 221, pp. 430
–
443, 2019, doi
:
10.1016/
j
.r
s
e
.2018.11.032.
[
11]
A
.
K
a
r
una
ni
t
hi
,
“
s
m
ut
_di
s
e
a
s
e
s
,”
F
i
gs
ha
r
e
.
A
c
c
e
s
s
e
d
:
M
a
r
.
13,
2023
.
[
O
nl
i
ne
.]
A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
f
i
gs
ha
r
e
.c
om
/
a
r
t
i
c
l
e
s
/
f
i
gur
e
/
s
m
ut
_di
s
e
a
s
e
s
/
21975236/
1.
[
12]
A
.
K
a
r
una
ni
t
hi
,
“
B
l
a
s
t
D
i
s
e
a
s
e
_D
a
t
a
s
e
t
,”
F
i
gs
har
e
.
A
c
c
e
s
s
e
d
:
J
a
n.
30,
2023
.
[
O
nl
i
ne
.]
A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
doi
.or
g/
10.6084/
m
9.f
i
gs
ha
r
e
.21975287.v1.
[
13]
A
.
K
a
r
una
ni
t
hi
,
“
B
a
c
t
e
r
a
i
l
bl
i
ght
_D
a
t
a
s
e
t
,”
F
i
gs
har
e
.
A
c
c
e
s
s
e
d
:
J
a
n.
30,
2023)
.
[
O
nl
i
ne
.]
A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
doi
.or
g/
10.6084/
m
9.f
i
gs
ha
r
e
.21975239.v1.
[
14]
N
.
B
ha
r
a
ni
dha
r
a
n,
S
.
R
.
S
a
nna
s
i
C
ha
kr
a
va
r
t
hy,
H
.
R
a
j
a
gur
u,
V
.
V
i
no
t
h
K
u
m
a
r
,
T
.
R
.
M
a
he
s
h,
a
nd
S
.
G
ul
uw
a
di
,
“
M
ul
t
i
c
l
a
s
s
pa
ddy
di
s
e
a
s
e
de
t
e
c
t
i
on
us
i
ng
f
i
l
t
e
r
-
ba
s
e
d
f
e
a
t
ur
e
t
r
a
ns
f
or
m
a
t
i
on
t
e
c
hni
que
,”
I
E
E
E
A
c
c
e
s
s
,
vol
.
11,
pp.
109477
–
109487,
2023,
doi
:
10.1109/
A
C
C
E
S
S
.2023.3322587.
[
15]
S
.
L
a
m
ba
e
t
al
.
,
“
A
nove
l
f
i
ne
-
t
une
d
de
e
p
-
l
e
a
r
ni
ng
-
ba
s
e
d
m
ul
t
i
-
c
l
a
s
s
c
l
a
s
s
i
f
i
e
r
f
or
s
e
ve
r
i
t
y
of
pa
ddy
l
e
a
f
di
s
e
a
s
e
s
,
”
F
r
ont
i
e
r
s
i
n
P
l
ant
Sc
i
e
nc
e
, vol
. 14, 2023, doi
:
10.3389/
f
pl
s
.2023.1234067.
[
16]
M
.
M
i
s
hr
a
,
P
.
C
houdhur
y,
a
nd
B
.
P
a
t
i
,
“
M
odi
f
i
e
d
r
i
de
-
N
N
opt
i
m
i
z
e
r
f
or
t
h
e
I
oT
ba
s
e
d
pl
a
nt
di
s
e
a
s
e
de
t
e
c
t
i
on,”
J
our
nal
o
f
A
m
bi
e
nt
I
nt
e
l
l
i
ge
nc
e
and H
um
ani
z
e
d C
om
put
i
ng
, vol
. 12, no. 1, pp. 691
–
703, 2
021, doi
:
10.1007/
s
12652
-
020
-
02051
-
6.
[
17]
K
.
P
.
F
e
r
e
nt
i
nos
,
“
D
e
e
p
l
e
a
r
ni
ng
m
ode
l
s
f
or
pl
a
nt
di
s
e
a
s
e
de
t
e
c
t
i
on
a
nd
di
a
g
nos
i
s
,”
C
om
put
e
r
s
and
E
l
e
c
t
r
oni
c
s
i
n
A
gr
i
c
ul
t
ur
e
,
vol
. 145, pp. 311
–
318, 2018, doi
:
10.1016/
j
.c
om
pa
g.2018.01.009.
[
18]
M
.
B
r
a
hi
m
i
,
K
.
B
oukha
l
f
a
,
a
nd
A
.
M
ous
s
a
oui
,
“
D
e
e
p
l
e
a
r
ni
ng
f
or
t
om
a
t
o
di
s
e
a
s
e
s
:
c
l
a
s
s
i
f
i
c
a
t
i
on
a
nd
s
ym
pt
om
s
vi
s
ua
l
i
z
a
t
i
on,
”
A
ppl
i
e
d A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
, vol
. 31, no. 4, pp. 299
–
315, 2017, doi
:
10.1080/
08839514.2017.1315516.
[
19]
J
.
A
ndr
e
w
,
J
.
E
uni
c
e
,
D
.
E
.
P
ope
s
c
u,
M
.
K
.
C
how
da
r
y,
a
nd
J
.
H
e
m
a
nt
h,
“
D
e
e
p
l
e
a
r
ni
ng
-
ba
s
e
d
l
e
a
f
di
s
e
a
s
e
de
t
e
c
t
i
on
i
n
c
r
ops
us
i
ng i
m
a
ge
s
f
or
a
gr
i
c
ul
t
ur
a
l
a
ppl
i
c
a
t
i
ons
,”
A
gr
onom
y
, vol
. 12, no. 10, 2022, d
oi
:
10.3390/
a
gr
onom
y12102395.
[
20]
R
.
M
.
J
.
A
l
-
A
kka
m
a
nd
M
.
S
.
M
.
A
l
t
a
e
i
,
“
P
l
a
nt
s
l
e
a
f
di
s
e
a
s
e
s
de
t
e
c
t
i
on
u
s
i
ng
de
e
p
l
e
a
r
ni
ng,”
I
r
aqi
J
our
nal
of
Sc
i
e
nc
e
,
vol
.
63,
no. 2, pp. 801
–
816, 2022, doi
:
10.24996/
i
j
s
.2022.63.2.34.
[
21]
N
.
K
um
a
r
,
N
a
ga
r
a
t
hna
,
a
nd
F
.
F
l
a
m
m
i
ni
,
“
Y
O
L
O
-
ba
s
e
d
l
i
ght
-
w
e
i
ght
de
e
p
l
e
a
r
ni
ng
m
ode
l
s
f
or
i
ns
e
c
t
d
e
t
e
c
t
i
on
s
ys
t
e
m
w
i
t
h
f
i
e
l
d
a
da
pt
i
on,”
A
gr
i
c
ul
t
ur
e
(
Sw
i
t
z
e
r
l
and)
, vol
. 13, no. 3, 2023, doi
:
10.3390/
a
gr
i
c
ul
t
ur
e
13030741.
[
22]
D
.
M
oha
n
a
nd
M
.
G
opa
l
R
a
j
,
“
Q
ua
l
i
t
y
a
na
l
ys
i
s
of
r
i
c
e
gr
a
i
ns
us
i
ng
A
N
N
a
nd
S
V
M
,”
J
our
nal
of
C
r
i
t
i
c
al
R
e
v
i
e
w
s
,
vol
.
7,
no.
1,
pp. 395
–
402, 2020, doi
:
10.31838/
j
c
r
.07.01.79.
[
23]
M
.
T
.
A
ha
d,
Y
.
L
i
,
B
.
S
ong,
a
nd
T
.
B
hui
ya
n,
“
C
om
pa
r
i
s
on
of
C
N
N
-
ba
s
e
d
de
e
p
l
e
a
r
ni
ng
a
r
c
hi
t
e
c
t
ur
e
s
f
or
r
i
c
e
di
s
e
a
s
e
s
c
l
a
s
s
i
f
i
c
a
t
i
on,”
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
i
n A
gr
i
c
ul
t
ur
e
, vol
. 9, pp. 22
–
35, 2023, do
i
:
10.1016/
j
.a
i
i
a
.2023.07.001.
[
24]
G
.
R
.
I
.
L
.
J
a
ya
s
oor
i
ya
a
nd
S
.
M
.
A
r
a
c
hc
hi
,
“
D
i
a
gnos
i
s
of
ba
c
t
e
r
i
a
l
l
e
a
f
b
l
i
g
ht
,
br
ow
n
s
pot
s
,
a
nd
l
e
a
f
s
m
ut
r
i
c
e
pl
a
nt
di
s
e
a
s
e
s
us
i
ng
l
i
ght
G
B
M
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
C
om
put
e
r
A
ppl
i
c
at
i
ons
,
vol
.
183,
no.
48,
pp.
53
–
58,
2022,
doi
:
10.5120/
i
j
c
a
2022921895.
[
25]
D
.
E
l
a
va
r
a
s
a
n
a
nd
P
.
M
.
D
.
V
i
nc
e
nt
,
“
C
r
op
yi
e
l
d
pr
e
di
c
t
i
on
us
i
ng
de
e
p
r
e
i
nf
or
c
e
m
e
nt
l
e
a
r
ni
ng
m
ode
l
f
o
r
s
us
t
a
i
na
bl
e
a
gr
a
r
i
a
n
a
ppl
i
c
a
t
i
ons
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 8, pp. 86886
–
86901, 2020, doi
:
10.1109/
A
C
C
E
S
S
.2020.2992480.
[
26]
V
.
B
a
dr
i
na
r
a
ya
na
n,
A
.
K
e
nda
l
l
,
a
nd
R
.
C
i
pol
l
a
,
“
S
e
gN
e
t
:
a
de
e
p
c
onvo
l
ut
i
ona
l
e
nc
ode
r
-
de
c
ode
r
a
r
c
hi
t
e
c
t
ur
e
f
or
i
m
a
ge
s
e
gm
e
nt
a
t
i
on,”
I
E
E
E
T
r
ans
ac
t
i
ons
on
P
at
t
e
r
n
A
nal
y
s
i
s
and
M
ac
hi
ne
I
nt
e
l
l
i
ge
nc
e
,
vol
.
39,
no.
12,
pp.
2481
–
2495,
2017,
doi
:
10.1109/
T
P
A
M
I
.2016.2644615.
[
27]
W
.
J
.
H
u,
J
.
F
a
n,
Y
.
X
.
D
u,
B
.
S
.
L
i
,
N
.
X
i
ong,
a
nd
E
.
B
e
kke
r
i
ng,
“
M
D
F
C
-
R
e
s
N
e
t
:
a
n
a
gr
i
c
ul
t
ur
a
l
I
oT
s
ys
t
e
m
t
o
a
c
c
ur
a
t
e
l
y
r
e
c
ogni
z
e
c
r
op di
s
e
a
s
e
s
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 8, pp. 115287
–
115298, 2020, doi
:
10.1109/
A
C
C
E
S
S
.2020.3001237.
[
28]
A
. S
t
e
phe
n, A
. P
uni
t
ha
, a
nd A
. C
h
a
ndr
a
s
e
ka
r
, “
O
pt
i
m
a
l
d
e
e
p ge
ne
r
a
t
i
ve
a
dv
e
r
s
a
r
i
a
l
ne
t
w
or
k a
nd c
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k f
o
r
r
i
c
e
l
e
a
f
di
s
e
a
s
e
pr
e
di
c
t
i
on,”
V
i
s
ual
C
om
put
e
r
, vol
. 40, no. 2, pp. 919
–
936, 202
4, doi
:
10.1007/
s
00371
-
023
-
02823
-
z.
Evaluation Warning : The document was created with Spire.PDF for Python.