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.
3554
~
3562
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
14
.i
5
.pp
3554
-
3562
3554
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
E
n
h
a
n
c
e
d
sol
ar
p
an
e
l
s f
au
l
t
d
e
t
e
c
t
i
on
ap
p
r
oac
h
u
s
i
n
g
l
i
gh
t
w
e
i
gh
t
Y
O
L
O
N
ai
m
a E
l
Y
an
b
oi
y
1
, M
oh
am
e
d
K
h
al
a
1
, I
s
m
ai
l
E
la
b
b
as
s
i
1
, N
ou
r
d
d
in
e
E
lh
aj
r
at
1
, O
m
ar
E
lo
u
t
as
s
i
2
,
Y
ou
s
s
e
f
E
l
H
as
s
ou
an
i
3
, C
h
ou
k
r
i
M
e
s
s
aou
d
i
1
1
D
e
pa
r
t
m
e
nt
of
O
pt
oe
l
e
c
t
r
oni
c
s
a
nd A
ppl
i
e
d E
ne
r
gy T
e
c
hni
que
s
, F
a
c
ul
t
y of
S
c
i
e
nc
e
a
nd T
e
c
hnol
ogy, M
oul
a
y I
s
m
a
i
l
U
ni
ve
r
s
i
t
y,
E
r
r
a
c
hi
di
a
, M
or
oc
c
o
2
M
a
t
e
r
i
a
l
s
a
nd
M
ode
l
l
i
ng L
a
bor
a
t
or
y, D
e
pa
r
t
m
e
nt
of
P
hys
i
c
s
, F
a
c
ul
t
y of
S
c
i
e
n
c
e
, M
oul
a
y I
s
m
a
i
l
U
ni
ve
r
s
i
t
y, M
e
kne
s
,
M
or
oc
c
o
3
D
e
pa
r
t
m
e
nt
of
N
e
w
E
ne
r
gi
e
s
a
nd M
a
t
e
r
i
a
l
s
E
ngi
ne
e
r
i
ng, F
a
c
ul
t
y of
S
c
i
e
nc
e
a
nd T
e
c
hnol
ogy, M
oul
a
y I
s
m
a
i
l
U
ni
ve
r
s
i
t
y,
E
r
r
a
c
hi
di
a
, M
or
oc
c
o
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
M
a
y 29
,
2024
R
e
vi
s
e
d
J
ul
20
,
2025
A
c
c
e
pt
e
d
A
ug
0
6
,
2025
Artificial
intelligence
(AI)
-
driven
fault
detection
improves
the
reliabi
lity
of
solar
energy
by
reducing
the
chances
of
system
failures.
However,
e
xisting
single
-
stage
object
detection
methods
excel
in
accu
racy
but
deman
d
high
computat
ional
resources,
preventin
g
seamless
integrat
ion
into
em
bedded
systems.
This
paper
introduces
a
lightweight
approach
using
YOLOv5,
which
incorporates
a
multi
-
backbone
design,
specifically
tailor
ed
for
accurate
fault
detection
in
solar
cells.
It
evaluates
YOLOv
5
and
TinyYOLOv5.
The
findings
emphasize
the
effectiveness
of
YOLOv
5l
with
Ghost
backbone,
particularly
notable
for
its
precision
rates
of
96%
for
faulty
and
93%
for
non
-
faulty
instances.
Additionally,
it
showcases
commendable
mean
average
precision
(mAP)
scores,
achieving
78%
at
an
intersectio
n
over
union
(
IoU)
threshold
of
0.5
and
72%
at
an
IoU
of
0.95.
Additi
onally,
YOLOv5_Ghost
emerges
as
the
optimal
selection,
showcasing
comp
etitive
precision,
processing
speed
of
42.1
giga
floating
point
operations
per second
(GFLOPS)
,
and
remarkable
efficiency
with
2.4
m
illion
parameters
.
This
evaluatio
n
underscores
the
effectivenes
s
of
YOLOv5
models,
t
hereby
leading to a
dvanced
solar ene
rgy techn
ology significa
ntly.
K
e
y
w
o
r
d
s
:
D
e
e
p l
e
a
r
ni
ng
F
a
ul
ts
de
te
c
ti
on
L
ig
ht
Y
O
L
O
P
hot
ovol
ta
ic
s
ys
te
m
S
m
a
r
t
de
te
c
ti
on
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
N
a
im
a
E
l
Y
a
nboi
y
D
e
pa
r
tm
e
nt
of
O
pt
oe
le
c
tr
oni
c
s
a
nd A
ppl
ie
d E
ne
r
gy T
e
c
hni
que
s
, F
a
c
ul
ty
of
S
c
ie
nc
e
a
nd T
e
c
hnol
ogy
M
oul
a
y I
s
m
a
il
U
ni
ve
r
s
it
y
E
r
r
a
c
hi
di
a
, M
or
oc
c
o
E
m
a
il
:
na
im
a
.e
ly
a
nboi
y@
gm
a
il
.c
om
1.
I
N
T
R
O
D
U
C
T
I
O
N
I
n
r
e
c
e
nt
ye
a
r
s
,
s
ol
a
r
pow
e
r
ha
s
s
ur
ge
d
a
s
a
pr
im
a
r
y
r
e
ne
w
a
bl
e
e
ne
r
gy
s
our
c
e
,
a
tt
r
a
c
ti
ng
gl
oba
l
a
tt
e
nt
io
n
a
nd
in
ve
s
tm
e
nt
.
T
hi
s
s
hi
f
t
e
m
pha
s
iz
e
s
it
s
pot
e
nt
ia
l
to
pr
ovi
de
e
ne
r
gy
in
de
pe
nde
nc
e
.
S
ol
a
r
pow
e
r
'
s
s
ig
ni
f
ic
a
nc
e
is
pa
r
ti
c
ul
a
r
ly
not
a
bl
e
in
a
ddr
e
s
s
in
g
gl
oba
l
e
ne
r
g
y
de
m
a
nds
w
hi
le
r
e
duc
in
g
r
e
li
a
nc
e
on
f
os
s
il
f
ue
ls
[
1]
.
I
nde
e
d,
it
c
a
n
s
ig
ni
f
ic
a
nt
ly
m
it
ig
a
te
gr
e
e
nhou
s
e
ga
s
e
m
is
s
io
ns
a
nd
c
om
ba
t
c
li
m
a
te
c
ha
ng
e
.
T
hi
s
i
s
e
s
pe
c
ia
ll
y
c
r
uc
i
a
l
in
a
r
e
a
s
m
os
t
a
f
f
e
c
te
d
by
e
nvi
r
onm
e
nt
a
l
de
g
r
a
da
ti
on.
M
or
e
ove
r
,
th
e
e
c
onomi
c
b
e
ne
f
it
s
of
s
ol
a
r
pow
e
r
a
r
e
s
ubs
ta
nt
ia
l.
T
he
s
ol
a
r
in
dus
tr
y
is
e
xpe
r
ie
nc
in
g
r
a
pi
d
gr
ow
th
[
2]
.
A
s
th
e
c
os
t
of
s
ol
a
r
pa
ne
ls
c
ont
in
ue
s
to
de
c
li
ne
a
nd
gov
e
r
nm
e
nt
in
c
e
nt
iv
e
s
e
n
c
our
a
ge
in
ve
s
tm
e
nt
in
s
ol
a
r
in
f
r
a
s
tr
uc
tu
r
e
,
th
e
e
c
onomi
c
f
e
a
s
ib
il
it
y
of
s
ol
a
r
po
w
e
r
be
c
om
e
s
in
c
r
e
a
s
in
gl
y
e
vi
de
nt
.
H
ow
e
ve
r
,
th
e
e
f
f
e
c
ti
ve
ne
s
s
of
s
ol
a
r
e
ne
r
gy
s
ys
te
m
s
f
a
c
e
s
s
ig
ni
f
ic
a
nt
c
ha
ll
e
ng
e
s
due
to
pot
e
nt
ia
l
f
a
ul
ts
th
a
t
c
a
n
oc
c
ur
dur
in
g
th
e
m
a
nuf
a
c
tu
r
in
g
or
ope
r
a
ti
on
of
s
ol
a
r
c
e
ll
s
[
3]
.
V
a
r
io
us
f
a
ul
ts
,
in
c
lu
di
ng
m
ic
r
oc
r
a
c
ks
,
hot
s
pot
s
,
s
oi
li
ng,
s
ha
dow
in
g,
a
nd
bi
r
d
dr
oppi
ngs
,
pos
e
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
E
nhanc
e
d s
ol
ar
pane
ls
f
aul
t
de
te
c
ti
on appr
oac
h u
s
in
g l
ig
ht
w
e
i
ght
Y
O
L
O
…
(
N
ai
m
a E
l
Y
anboiy
)
3555
c
r
it
ic
a
l
di
f
f
ic
ul
ty
to
th
e
e
f
f
ic
ie
nc
y
of
s
ol
a
r
e
ne
r
gy
s
ys
te
m
s
w
it
hi
n
th
e
phot
ovol
ta
ic
(
P
V
)
s
ys
te
m
[
4]
.
A
ddr
e
s
s
in
g
th
e
s
e
f
a
ul
ts
is
pa
r
a
m
ount
f
or
im
pr
ovi
ng
th
e
e
f
f
ic
ie
nc
y
of
P
V
ge
ne
r
a
ti
on
[
5]
.
C
ons
e
que
nt
ly
,
th
e
de
ve
lo
pm
e
nt
of
m
e
th
ods
f
or
s
m
a
r
t
de
te
c
ti
ng
f
a
ul
ts
in
s
ol
a
r
c
e
ll
s
hol
ds
s
ig
ni
f
ic
a
nt
im
por
ta
nc
e
[
6]
.
I
n
th
e
li
te
r
a
tu
r
e
,
va
r
io
us
te
c
hni
que
s
ha
v
e
be
e
n
e
xpl
or
e
d
f
or
de
te
c
ti
ng
f
a
ul
ts
in
s
ol
a
r
c
e
ll
s
,
br
oa
dl
y
c
a
te
gor
iz
e
d
in
to
im
a
ge
pr
oc
e
s
s
in
g
te
c
hni
que
s
,
tr
a
di
ti
ona
l
m
e
th
ods
s
uc
h
a
s
v
is
ua
l
in
s
pe
c
ti
on
a
nd
I
-
V
c
ur
ve
tr
a
c
in
g,
a
nd
a
r
ti
f
ic
ia
l
ne
ur
a
l
ne
twor
ks
(
A
N
N
s
)
[
7]
.
E
a
c
h
of
th
e
s
e
a
ppr
oa
c
he
s
ha
s
it
s
o
w
n
s
e
t
of
li
m
it
a
ti
ons
:
C
onve
nt
io
na
l
vi
s
ua
l
in
s
pe
c
ti
on
of
s
ol
a
r
c
e
ll
s
r
e
qui
r
e
s
s
pe
c
ia
li
z
e
d
e
qui
pm
e
n
t
a
nd
m
a
nua
l
e
xa
m
in
a
ti
on,
le
a
di
ng
to
la
bor
-
in
te
ns
iv
e
ta
s
ks
a
nd
s
ubj
e
c
ti
ve
out
c
om
e
s
[
8]
.
I
m
a
ge
pr
oc
e
s
s
in
g
te
c
hni
que
s
s
tr
uggl
e
w
it
h
c
om
pl
e
x
f
a
ul
ts
a
nd
e
nvi
r
onm
e
nt
a
l
c
ha
nge
s
,
pr
im
a
r
il
y
de
te
c
ti
ng
s
ur
f
a
c
e
-
le
ve
l
is
s
ue
s
.
I
nf
r
a
r
e
d
a
nd
E
le
c
tr
o
lu
m
in
e
s
c
e
nc
e
im
a
gi
ng,
th
ough
e
f
f
e
c
ti
ve
,
a
r
e
c
os
tl
y
a
nd
r
e
qui
r
e
s
p
e
c
ia
li
z
e
d
tr
a
in
in
g,
m
a
in
ly
s
ui
ta
bl
e
f
or
s
ur
f
a
c
e
-
le
v
e
l
de
te
c
ti
on
[
9]
.
A
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
(
A
I
)
,
pa
r
t
ic
ul
a
r
ly
de
e
p
le
a
r
ni
ng
(
D
L
)
m
e
th
ods
,
is
a
pow
e
r
f
ul
a
ppr
oa
c
h
f
or
f
a
ul
t
de
te
c
ti
on
in
s
ol
a
r
c
e
ll
s
.
H
ow
e
ve
r
,
it
r
e
qui
r
e
s
s
ig
ni
f
ic
a
nt
a
m
ount
s
of
da
ta
a
nd
c
om
put
a
ti
ona
l
r
e
s
our
c
e
s
.
M
or
e
ove
r
, i
t
f
a
c
e
s
l
im
it
a
ti
ons
i
n e
nvi
r
onm
e
nt
s
w
he
r
e
da
t
a
i
s
s
c
a
r
c
e
or
r
e
s
our
c
e
s
a
r
e
c
on
s
tr
a
in
e
d
[
10]
.
L
e
v
e
r
a
gi
ng
A
I
,
pa
r
ti
c
ul
a
r
ly
D
L
,
is
e
s
s
e
nt
ia
l
f
o
r
i
m
p
r
o
vi
ng
th
e
pe
r
f
or
m
a
nc
e
a
n
d
d
ur
a
bi
li
ty
o
f
s
o
la
r
e
ne
r
gy
s
ys
te
m
s
b
y
e
na
bl
in
g
th
e
a
u
to
m
a
te
d
a
n
d
p
r
e
c
is
e
i
de
n
ti
f
ic
a
t
io
n
o
f
va
r
io
us
f
a
ul
ts
i
n
s
ol
a
r
pa
ne
ls
[
11
]
.
A
I
m
e
t
hods
of
f
e
r
a
n
e
f
f
ic
ie
nt
s
ol
ut
io
n
f
o
r
e
a
r
ly
f
a
u
lt
de
t
e
c
ti
on
,
c
a
pa
bl
e
o
f
a
na
ly
z
in
g
la
r
ge
da
t
a
s
e
ts
a
c
c
u
r
a
te
ly
a
nd
in
a
ti
m
e
ly
m
a
nn
e
r
[
1
2]
.
T
h
is
le
d
t
o
s
ig
ni
f
ic
a
n
t
r
e
s
e
a
r
c
h
e
f
f
o
r
ts
f
o
c
us
e
d
on
de
t
e
c
t
in
g
a
nom
a
l
ie
s
i
n
PV
s
ys
te
m
s
.
J
a
n
a
r
th
a
na
n
e
t
al
.
[
13
]
p
r
e
s
e
nt
e
d
a
m
e
t
hod
ol
ogy
in
th
e
i
r
s
t
ud
y
a
i
m
e
d
a
t
a
dva
nc
i
ng
t
h
e
de
ve
lo
pm
e
nt
o
f
r
e
s
il
ie
nt
f
uz
z
y
lo
g
ic
s
ys
te
m
s
(
F
L
S
s
)
a
n
d
A
N
N
s
f
o
r
P
V
f
a
u
lt
de
te
c
ti
o
n
.
T
he
ir
r
e
s
e
a
r
c
h
h
ig
h
l
ig
h
ts
th
e
e
f
f
e
c
ti
ve
n
e
s
s
o
f
f
a
ul
t
i
de
n
ti
f
ic
a
ti
o
n
a
pp
r
oa
c
he
s
in
a
c
c
u
r
a
te
l
y
id
e
nt
if
yi
ng
di
s
ti
nc
t
f
a
u
lt
c
a
te
g
or
ie
s
,
in
c
lu
di
n
g
i
m
pa
ir
e
d
P
V
m
od
ul
e
s
a
nd
pa
r
t
i
a
l
s
ha
d
in
g
of
P
V
un
it
s
.
A
k
r
a
m
e
t
al
.
[
14
]
c
ond
uc
t
e
d
r
e
s
e
a
r
c
h
on
a
u
to
m
a
ti
n
g
t
he
de
te
c
ti
on
o
f
de
f
e
c
ts
i
n
P
V
m
od
ul
e
s
us
i
ng
i
n
f
r
a
r
e
d
i
m
a
ge
s
.
T
he
ir
s
tu
d
y
e
m
p
lo
y
e
d
is
o
la
te
d
D
L
a
nd
de
v
e
lo
p
-
m
o
de
l
tr
a
ns
f
e
r
l
e
a
r
n
in
g
te
c
h
ni
que
s
.
T
he
y
a
c
hi
e
ve
d
a
h
ig
h
a
ve
r
a
ge
a
c
c
ur
a
c
y
o
f
9
8.6
7%
us
in
g
a
l
ig
h
t
C
N
N
a
r
c
h
it
e
c
t
ur
e
f
o
r
a
n
is
o
la
te
d
t
r
a
in
e
d
m
o
de
l.
A
d
di
ti
ona
ll
y,
th
e
y
ut
il
iz
e
d
t
r
a
ns
f
e
r
le
a
r
n
in
g
by
p
r
e
-
t
r
a
i
ni
ng
a
ba
s
e
m
od
e
l
on
e
le
c
tr
ol
um
in
e
s
c
e
nc
e
im
a
ge
s
o
f
P
V
c
e
l
ls
a
n
d
f
i
ne
-
tu
ni
ng
it
o
n
i
nf
r
a
r
e
d
i
m
a
ge
s
.
P
r
a
b
ha
k
a
r
a
n
e
t
al
.
[
15
]
in
t
r
odu
c
e
s
th
e
r
e
a
l
-
t
im
e
m
u
lt
i
va
r
ia
n
t
de
e
p
le
a
r
ni
ng
m
od
e
l
(
R
M
V
D
M
)
.
T
he
m
ode
l
de
m
ons
t
r
a
te
s
im
pr
ove
d
pe
r
f
or
m
a
nc
e
w
h
il
e
r
e
qu
ir
in
g
le
s
s
c
om
p
ut
a
t
io
n
a
l
ti
m
e
,
un
de
r
s
c
o
r
i
ng
i
ts
e
f
f
ic
ie
nc
y
a
n
d
pr
a
c
t
ic
a
l
a
pp
li
c
a
bi
li
ty
.
R
a
m
í
r
e
z
e
t
al
.
[
1
6]
in
t
r
o
duc
e
s
a
n
i
nn
ova
ti
ve
m
e
t
ho
d
f
o
r
m
o
ni
to
r
i
ng
P
V
pa
ne
l
c
o
nd
it
io
n
by
in
te
g
r
a
ti
n
g
a
r
a
d
io
m
e
t
r
i
c
s
e
ns
o
r
w
it
h
an
unm
a
nne
d
a
e
r
ia
l
ve
h
ic
l
e
(
U
A
V
)
.
T
hi
s
a
p
pr
oa
c
h
de
te
c
ts
va
r
i
o
us
f
a
u
lt
s
,
in
c
lu
di
ng
ho
t
s
po
ts
a
nd
f
a
u
lt
y
c
e
ll
s
,
w
it
h
c
om
m
e
n
da
b
le
a
c
c
u
r
a
c
y
,
a
dva
n
c
i
ng
f
a
u
l
t
de
te
c
ti
on
f
o
r
P
V
s
ys
t
e
m
s
.
H
a
n
e
t
a
l.
[
1
7
]
p
r
o
pos
e
a
c
ut
ti
n
g
-
e
dge
m
e
t
hod
f
or
de
t
e
c
t
in
g
s
ol
a
r
pa
ne
l
de
f
e
c
ts
us
in
g
th
e
r
m
a
l
i
m
a
g
in
g,
e
m
p
lo
y
in
g
p
r
i
nc
i
pa
l
c
om
po
ne
n
t
a
na
l
ys
is
(
P
C
A
)
a
nd
i
nde
pe
n
de
n
t
c
om
po
ne
n
t
a
na
l
ys
is
(
I
C
A
)
te
c
hn
iq
u
e
s
.
T
hi
s
f
a
c
i
li
ta
t
e
s
e
a
s
y
de
f
e
c
t
id
e
n
ti
f
ic
a
ti
o
n
w
i
th
o
ut
c
os
tl
y
e
le
c
t
r
ic
a
l
de
te
c
ti
on
c
i
r
c
u
it
r
y
,
r
e
duc
in
g
t
im
e
a
n
d
c
os
ts
a
s
s
oc
i
a
te
d
w
it
h
de
te
c
ti
on
p
r
oc
e
du
r
e
s
.
I
n
th
is
s
tu
dy,
w
e
in
tr
oduc
e
a
n
im
pr
ove
d
Y
O
L
O
de
te
c
ti
on
m
od
e
l
w
it
h
a
n
a
r
c
hi
te
c
tu
r
e
f
in
e
-
tu
ne
d
f
o
r
e
f
f
ic
ie
nt
a
nd
pr
e
c
is
e
f
a
ul
ts
de
te
c
ti
on
in
P
V
m
odul
e
s
.
T
he
c
ont
r
ib
ut
in
g
poi
nt
s
of
th
is
r
e
s
e
a
r
c
h
in
c
lu
de
,
e
m
pl
oyi
ng
a
r
a
nge
of
da
ta
a
ugm
e
nt
a
ti
on
m
e
th
ods
to
of
f
e
r
pr
a
c
ti
c
a
l
s
ugge
s
ti
ons
f
or
e
f
f
e
c
ti
ve
da
ta
a
ugm
e
nt
a
ti
on,
e
nha
nc
in
g
th
e
a
c
c
ur
a
c
y
of
th
e
tr
a
in
in
g
m
ode
ls
.
E
xpl
oi
ti
ng
th
e
be
ne
f
it
s
of
Y
O
L
O
v5,
w
e
in
tr
oduc
e
a
n
a
dopt
e
d
Y
O
L
O
v5
ne
twor
k
de
s
ig
ne
d
f
or
de
f
e
c
t
de
t
e
c
ti
on
in
P
V
pa
n
e
ls
. T
he
obj
e
c
ti
ve
is
to
c
r
e
a
t
e
a
n
a
ut
om
a
te
d
de
te
c
ti
on
s
ys
t
e
m
th
a
t
e
xc
e
ls
in
a
c
c
ur
a
c
y,
c
om
put
a
ti
ona
l
e
f
f
ic
ie
nc
y,
a
nd
m
ode
l
s
iz
e
c
om
pa
c
tn
e
s
s
.
I
nt
e
gr
a
ti
ng
a
m
odi
f
ie
d
Y
O
L
O
v5
ti
ny
m
ode
l
by
s
ubs
ti
tu
ti
ng
th
e
or
ig
in
a
l
ba
c
kbone
w
it
h
ghos
t,
M
O
B
I
L
E
N
E
T
,
pr
e
-
pr
oc
e
s
s
in
g
a
nd
lo
c
a
li
z
a
ti
on
c
ont
r
ol
(
P
P
L
C
)
,
S
H
U
F
F
L
E
,
a
nd
Y
O
L
O
v5l
E
f
f
ic
ie
nt
L
it
a
r
c
hi
te
c
tu
r
e
s
.
T
he
r
e
s
ul
ts
of
th
e
pr
opos
e
d
a
ppr
oa
c
h
de
m
on
s
tr
a
te
th
a
t
th
e
Y
O
L
O
v5G
hos
t
-
li
ght
w
e
ig
ht
m
ode
l
s
uc
c
e
s
s
f
ul
ly
de
te
c
t
s
f
a
ul
ts
in
P
V
s
ys
t
e
m
s
,
a
c
hi
e
vi
ng
th
e
hi
gh
e
s
t
a
ve
r
a
ge
pr
e
c
is
io
n
of
95%
,
out
pe
r
f
or
m
in
g
Y
O
L
O
v5s
w
hi
c
h r
e
a
c
he
d 74.8%
.
T
he
r
e
s
t
of
th
e
doc
um
e
nt
is
s
tr
uc
tu
r
e
d
a
s
f
ol
lo
w
s
:
s
e
c
ti
on
2
out
li
ne
s
th
e
Y
O
L
O
m
ode
ls
w
e
pr
opos
e
.
S
e
c
ti
on
3
c
ont
a
in
s
th
e
pr
e
s
e
nt
a
ti
on
a
nd
di
s
c
us
s
io
n
of
th
e
e
xpe
r
im
e
nt
a
l
r
e
s
ul
ts
.
L
a
s
tl
y,
in
s
e
c
ti
on
4,
w
e
d
e
lv
e
in
to
t
he
c
onc
lu
s
io
n a
nd f
ut
ur
e
w
or
k.
2.
M
E
T
H
O
D
T
he
pr
oc
e
s
s
be
gi
ns
w
it
h
da
ta
a
c
qui
s
it
io
n,
a
s
s
how
n
in
th
e
F
ig
ur
e
1
de
s
c
r
ib
in
g
th
e
a
r
c
hi
te
c
tu
r
a
l
bl
ue
pr
in
t
of
th
e
p
r
opos
e
d
m
e
th
odol
ogy,
w
he
r
e
a
va
r
ie
ty
of
da
t
a
a
r
e
c
ol
le
c
te
d
f
r
om
s
ol
a
r
pa
ne
l
in
s
ta
ll
a
ti
ons
.
T
hi
s
da
ta
is
th
e
n
pr
oc
e
s
s
e
d
a
nd
pr
e
-
pr
oc
e
s
s
e
d
to
e
n
s
ur
e
it
s
qua
li
ty
a
nd
s
ui
ta
bi
li
ty
f
or
tr
a
in
in
g.
S
ubs
e
que
nt
ly
,
th
e
m
ode
ls
a
r
e
tr
a
in
e
d
us
in
g
a
dva
nc
e
d
a
lg
or
it
hm
s
,
le
ve
r
a
gi
ng
te
c
hni
que
s
s
uc
h
a
s
D
L
a
nd
pa
tt
e
r
n
r
e
c
ogni
ti
on
to
de
te
c
t
s
ubt
le
f
a
ul
ts
w
it
hi
n
th
e
s
ol
a
r
pa
ne
ls
.
T
hr
ough
it
e
r
a
ti
ve
tr
a
in
in
g,
th
e
m
ode
ls
le
a
r
n
to
id
e
nt
if
y
a
nom
a
li
e
s
in
di
c
a
ti
ve
of
pot
e
nt
ia
l
f
a
ul
ts
.
T
he
ul
ti
m
a
te
out
put
of
th
is
c
om
pr
e
he
ns
iv
e
a
ppr
oa
c
h
is
a
s
m
a
r
t
f
a
ul
t
de
te
c
ti
on s
ys
te
m
c
a
pa
bl
e
of
a
c
c
ur
a
te
ly
i
de
nt
if
yi
ng i
s
s
ue
s
w
it
hi
n s
ol
a
r
pa
ne
l
a
r
r
a
ys
.
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
:
3554
-
3562
3556
F
ig
ur
e
1. C
om
pr
e
he
ns
iv
e
ove
r
vi
e
w
:
a
r
c
hi
te
c
tu
r
a
l
bl
ue
pr
in
t
of
t
he
pr
opos
e
d m
e
th
odol
ogy
I
n
th
is
r
e
s
e
a
r
c
h,
da
ta
a
c
qui
s
it
io
n
in
vol
ve
d
ut
il
iz
in
g
a
d
a
ta
ba
s
e
f
r
om
th
r
e
e
PV
m
odul
e
te
c
hnol
ogi
e
s
m
onoc
r
ys
ta
ll
in
e
(
m
-
S
i
)
,
pol
yc
r
ys
ta
ll
in
e
(
p
-
S
i)
a
nd
a
m
or
pha
(a
-
S
i)
in
E
r
r
a
c
hi
di
a
,
M
or
oc
c
o
a
s
s
how
n
in
F
ig
ur
e
2.
I
m
a
ge
s
c
a
pt
ur
e
d
w
it
h
a
hi
gh
-
r
e
s
ol
ut
io
n
c
a
m
e
r
a
de
pi
c
te
d
va
r
io
us
a
nom
a
li
e
s
li
ke
dus
t
a
c
c
um
ul
a
ti
on,
s
ha
di
ng,
c
r
a
c
ks
,
a
nd
bi
r
d
dr
oppi
ngs
a
s
il
lu
s
tr
a
te
d
in
F
ig
ur
e
3
.
T
he
da
ta
s
e
t
w
a
s
e
xpa
nd
e
d
to
6
,
300
im
a
ge
s
us
in
g a
ugm
e
nt
a
ti
on t
e
c
hni
que
s
, w
it
h 80%
a
ll
oc
a
te
d f
or
t
r
a
in
in
g
a
nd 20%
f
or
t
e
s
ti
ng. T
hi
s
da
ta
s
e
t
w
a
s
m
e
r
ge
d
w
it
h
a
not
he
r
R
obof
lo
w
da
ta
s
e
t
to
im
pr
ove
tr
a
in
in
g
[
18]
.
A
ugm
e
nt
a
ti
on
te
c
hni
que
s
in
c
lu
d
e
d
hor
iz
ont
a
l
f
li
ppi
ng, c
r
oppi
ng w
it
h z
oom
(
0%
-
20%
)
, a
nd b
r
ig
ht
ne
s
s
va
r
ia
ti
on (
-
25%
t
o +
25%
)
.
C
r
a
c
ks
B
i
r
d dr
oppi
ngs
D
us
t
a
c
c
um
ul
a
t
i
on
S
ha
di
ng
F
ig
ur
e
3.
A
n a
r
r
a
y of
s
ol
a
r
pa
ne
l
f
a
ul
ts
:
a
vi
s
ua
l
gui
de
t
o c
om
m
on i
s
s
ue
s
2
.1.
Y
O
L
O
:
a
lg
or
it
h
m
s
an
d
ar
c
h
it
e
c
t
u
r
al
f
r
am
e
w
or
k
s
T
he
Y
O
L
O
a
r
c
hi
te
c
tu
r
e
c
ons
i
s
ts
of
th
r
e
e
m
a
in
c
om
pone
nt
s
:
th
e
ba
c
kbone
,
ne
c
k,
a
nd
h
e
a
d
[
19]
.
T
he
s
e
e
le
m
e
nt
s
,
w
hi
c
h
m
a
y
va
r
y
a
c
r
os
s
di
f
f
e
r
e
nt
Y
O
L
O
ve
r
s
io
ns
,
pl
a
y
c
r
uc
ia
l
r
ol
e
s
in
de
te
r
m
in
in
g
th
e
m
ode
l'
s
s
pe
e
d
a
nd
a
c
c
ur
a
c
y
[
20]
.
T
hi
s
s
ub
s
e
c
ti
on
of
f
e
r
s
a
n
in
s
i
ght
in
to
th
e
ne
twor
k
a
r
c
hi
te
c
tu
r
e
of
Y
O
L
O
v5,
w
e
ll
-
known
f
or
it
s
a
dva
nc
e
d
de
te
c
ti
on
c
a
pa
bi
li
ti
e
s
a
c
r
os
s
di
ve
r
s
e
s
c
a
le
s
[
21]
.
T
he
c
or
e
a
r
c
hi
te
c
tu
r
e
,
de
pi
c
te
d
in
F
ig
ur
e
4,
r
e
li
e
s
on
a
ba
c
kbone
s
e
r
vi
ng
a
s
a
f
e
a
tu
r
e
e
xt
r
a
c
t
or
,
e
m
pl
oyi
ng
a
c
onvolut
io
na
l
ne
ur
a
l
ne
twor
k
(
C
N
N
)
tr
a
in
e
d
on
e
xt
e
ns
iv
e
da
ta
s
e
ts
s
uc
h
a
s
I
m
a
g
e
N
e
t.
Y
O
L
O
v5
ut
il
iz
e
s
th
e
C
S
P
D
a
r
kne
t5
3
b
a
c
kbone
,
c
hos
e
n
f
or
it
s
e
f
f
e
c
ti
ve
ne
s
s
in
c
a
pt
ur
in
g
f
e
a
tu
r
e
s
f
r
om
in
p
ut
im
a
ge
s
.
A
ddi
ti
ona
ll
y,
Y
O
L
O
v5
in
te
gr
a
te
s
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
E
nhanc
e
d s
ol
ar
pane
ls
f
aul
t
de
te
c
ti
on appr
oac
h u
s
in
g l
ig
ht
w
e
i
ght
Y
O
L
O
…
(
N
ai
m
a E
l
Y
anboiy
)
3557
te
c
hni
que
s
li
ke
f
e
a
tu
r
e
pyr
a
m
id
ne
twor
k
(
F
P
N
)
a
nd
pa
th
a
ggr
e
ga
ti
on
ne
twor
k
(
P
A
N
)
[
22
]
.
F
P
N
in
v
o
l
v
e
s
up
-
s
a
m
p
li
n
g
t
h
e
ou
t
p
ut
f
e
a
t
ur
e
m
a
p
(
C
3,
C
4
,
a
n
d
C
5
)
f
r
o
m
v
a
r
i
ous
c
o
n
v
o
l
u
t
io
n
a
l
d
o
w
n
-
s
a
m
pl
i
n
g
o
p
e
r
a
ti
o
n
s
[
2
3]
,
g
e
n
e
r
a
t
in
g
m
u
l
t
ip
l
e
n
e
w
f
e
a
tu
r
e
m
a
p
s
(
P
3
,
P
4
,
a
nd
P
5)
to
e
n
h
a
n
c
e
t
a
r
g
e
t
d
e
t
e
c
t
io
n
a
c
r
o
s
s
a
v
a
r
i
e
t
y
of
s
c
a
l
e
s
[
2
4]
.
F
ig
ur
e
4.
Y
O
L
O
v5 a
r
c
hi
te
c
tu
r
e
T
hi
s
pa
pe
r
f
oc
us
e
s
on
L
ig
ht
w
e
ig
ht
Y
O
L
O
v5
m
ode
ls
,
a
im
in
g
to
e
nha
nc
e
de
f
e
c
t
de
te
c
ti
on
in
s
ol
a
r
pa
ne
ls
.
V
a
r
io
us
Y
O
L
O
v5
-
M
ul
ti
ba
c
kbone
m
ode
ls
w
e
r
e
u
ti
li
z
e
d
a
s
s
how
n
in
F
ig
ur
e
5,
in
c
lu
di
n
g
Y
O
L
O
v5l
E
f
f
ic
ie
nt
L
it
e
,
Y
O
L
O
v5l
G
hos
t,
Y
O
L
O
v5l
M
obi
le
ne
tv
3S
m
a
ll
,
Y
O
L
O
v5l
P
P
-
L
C
,
a
nd
Y
O
L
O
v5l
S
huf
f
le
.
Y
O
L
O
v5l
E
f
f
ic
ie
nt
L
it
e
in
te
gr
a
te
s
a
c
us
to
m
iz
e
d
E
f
f
ic
ie
nt
N
e
tL
it
e
ba
c
kbone
,
w
hi
le
Y
O
L
O
v5l
G
hos
t
f
e
a
tu
r
e
s
a
G
hos
tn
e
t
-
ba
s
e
d
ba
c
kbone
f
or
m
ul
ti
-
s
c
a
le
f
e
a
tu
r
e
f
us
io
n.
Y
O
L
O
v5l
P
P
-
L
C
ut
il
iz
e
s
M
obi
le
N
e
tv
3S
m
a
ll
a
r
c
hi
te
c
tu
r
e
,
a
nd
Y
O
L
O
v5l
P
P
-
L
C
N
e
t
e
m
pl
oys
P
P
-
L
C
N
e
t
a
r
c
hi
te
c
tu
r
e
,
bot
h
e
nha
nc
in
g
obj
e
c
t
de
te
c
ti
on
c
a
pa
bi
li
ti
e
s
.
L
a
s
tl
y,
Y
O
L
O
v5l
S
huf
f
le
N
e
tV2
us
e
s
S
huf
f
le
N
e
tV2_I
nve
r
te
dR
e
s
id
ua
l
m
odul
e
s
f
or
f
e
a
tu
r
e
e
xt
r
a
c
ti
on, e
a
c
h w
it
h s
pe
c
if
ic
pa
r
a
m
e
te
r
s
t
a
il
or
e
d f
or
i
m
pr
ove
d de
te
c
ti
on a
c
c
ur
a
c
y.
F
ig
ur
e
5.
G
e
ne
r
a
l
f
r
a
m
e
w
or
k f
or
f
a
ul
t
de
te
c
ti
on i
n s
ol
a
r
pa
ne
ls
2
.2.
P
e
r
f
or
m
an
c
e
as
s
e
s
s
m
e
n
t
T
o
e
va
lu
a
te
th
e
e
f
f
e
c
ti
ve
ne
s
s
of
e
a
c
h
m
ode
l,
di
ve
r
s
e
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
in
c
lu
di
ng
a
c
c
ur
a
c
y,
pr
e
c
is
io
n, r
e
c
a
ll
, F
1
-
s
c
or
e
[
25]
, a
nd me
a
n a
ve
r
a
ge
pr
e
c
i
s
io
n (
m
A
P
)
a
r
e
c
a
lc
ul
a
te
d
[
26]
.
=
(
+
)
(
+
+
+
)
(
1
)
=
(
+
)
(
2
)
=
(
+
)
(
3
)
1
−
=
2
×
(
×
)
(
+
)
(
4
)
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
:
3554
-
3562
3558
=
1
∑
=
1
(
5
)
W
he
r
e
is
t
he
t
ot
a
l
num
be
r
of
c
la
s
s
e
s
a
nd
is
t
he
a
ve
r
a
g
e
pr
e
c
is
i
on f
or
c
la
s
s
.
3.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
I
n e
xa
m
in
in
g t
he
r
e
s
ul
ts
of
t
he
pr
opos
e
d m
ode
ls
f
or
f
a
ul
t
de
te
c
ti
on i
n s
ol
a
r
pa
ne
ls
, i
t
is
c
le
a
r
t
o f
oc
us
on
va
r
io
us
c
r
it
ic
a
l
a
s
p
e
c
ts
of
th
e
ir
pe
r
f
or
m
a
nc
e
.
A
m
ong
th
e
s
e
c
ons
id
e
r
a
ti
ons
,
c
om
put
a
ti
ona
l
c
o
s
ts
a
s
s
oc
ia
te
d
w
it
h
tr
a
in
in
g
e
a
c
h
va
r
ia
nt
e
m
e
r
ge
a
s
a
s
ig
ni
f
ic
a
nt
f
a
c
to
r
.
T
a
b
le
1
s
um
m
a
r
iz
e
s
th
e
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
of
Y
O
L
O
v5,
Y
O
L
O
v5
L
ig
ht
,
a
nd
Y
O
L
O
v8
m
ode
ls
on
th
e
va
li
da
ti
on
s
e
t.
Y
O
L
O
v5G
hos
t
a
c
hi
e
ve
s
th
e
hi
ghe
s
t
pr
e
c
is
io
n
of
0.95,
f
ol
lo
w
e
d
by
Y
O
L
O
v5
E
f
f
ic
ie
nt
N
e
t
w
it
h
0.9
2.
Y
O
L
O
v8
de
m
ons
tr
a
te
s
th
e
hi
ghe
s
t
r
e
c
a
ll
a
t
0.89,
w
hi
le
Y
O
L
O
v5
a
nd
Y
O
L
O
v5L
ig
ht
m
ode
ls
e
xhi
bi
t
r
e
c
a
ll
s
be
twe
e
n
0.61
a
nd
0.68.
I
n
te
r
m
s
of
m
A
P
@
50,
Y
O
L
O
v8
le
a
ds
w
it
h
0.94,
f
ol
lo
w
e
d
c
lo
s
e
ly
by
Y
O
L
O
v5s
w
it
h
0.85.
Y
O
L
O
_G
hos
t
e
xc
e
ls
in
pr
oc
e
s
s
in
g
s
p
e
e
d
a
t
42.1
g
ig
a
f
lo
a
ti
ng
poi
nt
ope
r
a
ti
ons
pe
r
s
e
c
ond
(
G
F
L
O
P
S
)
,
w
it
h
24,226,831
pa
r
a
m
e
te
r
s
.
D
e
s
pi
te
ha
vi
ng
f
e
w
e
r
pa
r
a
m
e
te
r
s
,
Y
O
L
O
v5G
hos
t
m
a
in
ta
in
s
c
om
pe
ti
ti
ve
pr
e
c
is
io
n
a
nd
pr
oc
e
s
s
in
g
s
pe
e
d,
m
a
ki
ng
it
a
s
tr
ong
c
hoi
c
e
.
Y
O
L
O
v5
G
hos
t
s
ta
nds
out
w
it
h
a
pr
e
c
is
io
n
va
lu
e
of
0.95,
s
how
c
a
s
in
g
it
s
a
bi
li
ty
to
m
in
im
iz
e
f
a
ls
e
pos
it
iv
e
de
te
c
ti
ons
de
f
e
c
ts
on
s
ol
a
r
pa
ne
ls
,
c
lo
s
e
ly
f
ol
lo
w
e
d
by
Y
O
L
O
v5E
f
f
ic
ie
nt
N
e
t
a
t
0.92.
Y
O
L
O
v8
e
xc
e
ls
in
r
e
c
a
ll
a
t
0.89.
H
ow
e
ve
r
,
Y
O
L
O
v5
a
nd
Y
O
L
O
v5L
ig
ht
m
ode
ls
s
how
lo
w
e
r
r
e
c
a
ll
v
a
lu
e
s
(
0.61
to
0.68)
,
s
ugge
s
ti
ng
li
m
it
a
ti
ons
in
c
a
pt
ur
in
g
tr
ue
pos
it
iv
e
in
s
ta
nc
e
s
.
Y
O
L
O
v8
le
a
d
s
in
m
A
P
@
50
w
it
h
0.94,
f
ol
lo
w
e
d
by
Y
O
L
O
v5s
a
t
0.85,
w
hi
le
Y
O
L
O
G
hos
t
a
m
ong
Y
O
L
O
L
ig
ht
m
ode
ls
s
c
or
e
s
0.77.
T
he
s
e
r
e
s
ul
ts
hi
ghl
ig
ht
va
r
io
us
s
tr
e
ngt
hs
a
nd t
r
a
de
-
of
f
s
a
c
r
os
s
m
ode
ls
. I
n t
e
r
m
s
of
pa
r
a
m
e
te
r
s
a
nd pr
oc
e
s
s
in
g s
pe
e
d,
Y
O
L
O
G
hos
t
ba
la
nc
e
s
w
e
ll
w
it
h
24,226,831
pa
r
a
m
e
te
r
s
a
nd
a
pr
oc
e
s
s
in
g
s
pe
e
d
of
42.1
G
F
L
O
P
S
.
D
e
s
pi
te
ope
r
a
ti
ng
w
it
h
m
or
e
pa
r
a
m
e
te
r
s
(
46,113,663
a
nd
43,608,150)
,
Y
O
L
O
v5
a
nd
Y
O
L
O
v8
e
xhi
bi
t
f
a
s
te
r
pr
oc
e
s
s
in
g
s
pe
e
ds
a
t
107.7
G
F
L
O
P
S
a
nd
164.8
G
F
L
O
P
S
.
T
hi
s
in
di
c
a
te
s
Y
O
L
O
_G
hos
t'
s
c
om
m
e
nda
bl
e
tr
a
de
-
of
f
be
twe
e
n pa
r
a
m
e
te
r
s
a
nd pr
oc
e
s
s
in
g e
f
f
ic
ie
nc
y.
F
ig
ur
e
6
il
lu
s
tr
a
te
s
th
e
e
xpe
r
im
e
nt
a
l
r
e
s
ul
ts
,
s
how
c
a
s
in
g
m
A
P
50
a
nd
m
A
P
95
pe
r
f
or
m
a
nc
e
s
.
Y
O
L
O
v8
de
m
ons
tr
a
te
d t
he
hi
ghe
s
t
va
lu
e
s
, f
ol
lo
w
e
d by YO
L
O
v5. I
n t
he
c
a
s
e
of
t
he
l
ig
ht
m
ode
l,
Y
O
L
O
G
hos
t
e
xhi
bi
te
d
th
e
be
s
t
pe
r
f
or
m
a
nc
e
,
w
hi
le
Y
O
L
O
S
huf
f
le
s
how
e
d
lo
w
e
r
va
lu
e
s
.
L
os
s
va
lu
e
s
f
or
Y
O
L
O
G
hos
t
a
nd
Y
O
L
O
E
f
f
ic
ie
nt
w
e
r
e
a
r
ound
0.006
a
s
s
how
n
in
F
ig
ur
e
7.
T
hi
s
pa
pe
r
e
m
pl
oys
di
ve
r
s
e
m
e
th
ods
to
e
nha
nc
e
ba
c
kbone
s
, e
xpl
a
in
in
g pe
r
f
or
m
a
nc
e
e
nha
nc
e
m
e
nt
s
a
nd pote
nt
ia
l
a
ppl
ic
a
ti
ons
.
T
a
bl
e
1
.
T
he
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
of
t
he
pr
opos
e
d m
ode
l
s
M
ode
l
ba
c
kbone
P
r
e
c
i
s
i
on
R
e
c
a
l
l
m
A
P
@
0.
05
m
A
P
@
0.
95
NP
M
ode
l
l
a
ye
r
s
C
P
U
time
P
r
oc
e
s
s
i
ng s
pe
e
d
(
G
F
L
O
P
S
)
Y
O
L
O
v5l
G
hos
t
95
63
77
70
24226831
552
45
.
5
42
.
1
Y
O
L
O
v5l
M
obi
l
e
ne
t
v3
85
60
75
67
20317981
337
40
.
5
38
.
2
Y
O
L
O
v5l
P
P
-
LC
90
60
75
67
21588991
294
44
.
2
41
.
5
Y
O
L
O
v5l
S
huf
f
l
e
87
64
76
68
21210447
361
43
.
2
40
.
4
E
f
f
i
c
i
e
nt
L
i
t
e
92
61
76
67
22958935
298
42
.
5
44
Y
O
L
O
v5l
96
68
84
79
46113663
267
48
.
4
107
.
7
Y
O
L
O
v3_t
i
ny
94
64
78
66
8669002
38
32
.
7
12
.
9
Y
O
L
O
v8
90
89
94
89
43608150
268
30
.
2
164
.
8
F
ig
ur
e
6. mAP
50
a
nd mAP
95
of
e
a
c
h m
ode
l
s
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
E
nhanc
e
d s
ol
ar
pane
ls
f
aul
t
de
te
c
ti
on appr
oac
h u
s
in
g l
ig
ht
w
e
i
ght
Y
O
L
O
…
(
N
ai
m
a E
l
Y
anboiy
)
3559
F
ig
ur
e
7.
P
r
e
c
is
io
n a
nd r
e
c
a
ll
of
t
he
m
ode
ls
T
he
pr
e
c
i
s
io
n
-
c
onf
id
e
nc
e
c
ur
ve
a
nd
r
e
c
a
ll
-
c
onf
id
e
nc
e
c
ur
ve
pr
om
in
e
nt
ly
e
xhi
bi
t
e
le
va
te
d
va
lu
e
s
,
s
pe
c
if
ic
a
ll
y
0.98
a
nd
0.96,
r
e
s
p
e
c
ti
ve
ly
,
unde
r
s
c
or
in
g
th
e
m
o
de
l'
s
e
f
f
ic
ie
nc
y
in
a
c
c
ur
a
te
ly
id
e
nt
if
yi
ng
bot
h
f
a
ul
ty
a
nd
non
-
f
a
ul
ty
s
ol
a
r
pa
ne
ls
.
T
he
s
e
m
e
tr
ic
s
r
e
f
le
c
t
a
r
obus
t
pe
r
f
or
m
a
nc
e
in
te
r
m
s
of
pr
e
c
is
io
n
a
nd
r
e
c
a
ll
,
c
r
uc
ia
l
f
or
e
ns
ur
in
g
a
r
e
li
a
bl
e
id
e
nt
if
ic
a
ti
on
p
r
oc
e
s
s
.
F
ur
th
e
r
m
or
e
,
th
e
pr
e
c
is
io
n
-
r
e
c
a
ll
va
lu
e
s
of
0.96
f
or
f
a
ul
ty
pa
ne
ls
a
nd
0.90
f
or
non
-
f
a
ul
ty
pa
ne
ls
,
a
s
il
lu
s
tr
a
te
d
in
F
ig
ur
e
8,
e
m
pha
s
iz
e
th
e
m
ode
l'
s
a
bi
li
ty
to
m
a
in
ta
in
hi
gh
pr
e
c
is
io
n
w
hi
le
e
f
f
e
c
ti
ve
ly
c
a
pt
ur
in
g
in
s
t
a
nc
e
s
of
bot
h
f
a
ul
ty
a
nd
non
-
f
a
ul
ty
s
ol
a
r
pa
n
e
ls
.
T
he
s
e
f
in
di
ngs
hi
ghl
ig
ht
th
e
m
ode
l'
s
ba
la
nc
e
d
pe
r
f
or
m
a
nc
e
i
n
a
c
hi
e
vi
ng
a
c
c
ur
a
te
de
te
c
ti
on
a
c
r
os
s
di
ve
r
s
e
s
c
e
na
r
io
s
,
c
ont
r
ib
ut
in
g
to
it
s
ove
r
a
ll
e
f
f
ic
a
c
y
in
s
ol
a
r
pa
ne
l
de
f
e
c
t
de
te
c
ti
on.
A
f
te
r
e
xa
m
in
in
g
th
e
va
li
da
ti
on
ba
tc
h
pr
e
di
c
ti
ons
,
it
be
c
om
e
s
e
vi
de
nt
th
a
t
th
e
m
ode
l
a
c
c
ur
a
te
l
y
id
e
nt
if
ie
s
a
nd
de
li
ne
a
te
s
de
f
e
c
ts
li
ke
c
r
a
c
k
s
,
s
oi
li
ng,
s
ha
dow
,
a
nd
bi
r
d
dr
opp
in
g,
a
s
de
pi
c
te
d
in
F
ig
ur
e
9
.
T
hi
s
unde
r
s
c
or
e
s
th
e
pr
of
ic
ie
nc
y
of
l
ig
ht
w
e
ig
ht
Y
O
L
O
m
ode
ls
i
n de
te
c
ti
ng a
nd c
a
t
e
gor
iz
in
g s
uc
h a
nom
a
li
e
s
.
F
ig
ur
e
8.
T
he
pr
e
c
is
io
n
-
c
onf
id
e
nc
e
c
ur
ve
a
nd r
e
c
a
ll
-
c
onf
id
e
nc
e
c
ur
ve
of
t
he
be
s
t
li
ght
m
ode
l
F
ig
ur
e
9. R
e
s
ul
ts
of
de
te
c
te
d f
a
ul
ts
i
n
s
ol
a
r
pa
ne
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
:
3554
-
3562
3560
4.
C
O
N
C
L
U
S
I
O
N
I
n
th
is
s
t
udy,
th
e
a
pp
li
c
a
ti
on
of
m
ul
ti
pl
e
Y
O
L
O
m
ode
l
s
f
or
de
f
e
c
t
d
e
te
c
ti
on
on
th
e
s
ur
f
a
c
e
s
of
s
ol
a
r
pa
ne
l
s
w
a
s
e
xp
lo
r
e
d.
C
on
s
id
e
r
in
g
t
he
s
e
t
of
5040
i
m
a
g
e
s
us
e
d
t
o
tr
a
in
th
e
m
od
e
l
s
,
th
e
r
e
s
e
a
r
c
h
r
e
s
ul
t
s
pr
o
vi
de
c
om
pe
l
li
ng
e
vi
d
e
nc
e
th
a
t
th
e
Y
O
L
O
v5G
os
t
-
li
g
ht
w
e
ig
ht
m
o
d
e
l
s
uc
c
e
s
s
f
ul
l
y
a
c
hi
e
ve
s
th
e
go
a
l
of
d
e
f
e
c
t
de
te
c
ti
on
in
P
V
s
y
s
te
m
s
,
w
it
h
th
e
hi
gh
e
s
t
a
ve
r
a
ge
pr
e
c
is
i
on
o
f
95%
c
om
pa
r
e
d
t
o
Y
O
L
O
v5
s
w
hi
c
h
r
e
a
c
h
e
d
74.8%
.
T
h
e
c
o
m
pa
r
a
ti
ve
r
e
s
ul
t
s
b
e
twe
e
n
Y
O
L
O
v5l
G
hos
t
a
nd
Y
O
L
O
v5l
hi
gh
li
ght
s
ig
n
if
ic
a
nt
di
f
f
e
r
e
nc
e
s
in
te
r
m
s
of
pr
e
c
i
s
io
n,
num
be
r
of
p
a
r
a
m
e
te
r
s
,
C
P
U
ti
m
e
,
a
nd
pr
oc
e
s
s
i
ng
s
pe
e
d.
W
hi
le
Y
O
L
O
v
5l
a
c
hi
e
ve
s
a
s
li
ght
l
y
hi
g
he
r
pr
e
c
i
s
io
n
of
ov
e
r
96%
,
Y
O
L
O
v5l
G
hos
t
a
tt
a
in
s
a
pr
e
c
i
s
io
n
of
95%
.
H
o
w
e
v
e
r
,
Y
O
L
O
v5l
G
ho
s
t
f
e
a
tu
r
e
s
a
li
ght
e
r
a
r
c
hi
te
c
tu
r
e
w
it
h
onl
y
24,
226,8
31
pa
r
a
m
e
t
e
r
s
c
om
pa
r
e
d
to
4
6,113
,663
f
or
Y
O
L
O
v5
l,
w
hi
c
h
c
a
n be
a
dv
a
nt
a
ge
o
us
f
or
e
m
be
dd
e
d s
y
s
te
m
s
w
it
h
li
m
it
e
d
r
e
s
our
c
e
s
.
F
ut
ur
e
d
ir
e
c
ti
ons
in
c
lu
d
e
e
xpl
or
in
g
a
m
or
e
c
onc
i
s
e
ba
c
kbo
ne
s
tr
uc
tu
r
e
a
nd
a
n
e
f
f
ic
ie
nt
f
e
a
tu
r
e
f
us
i
on me
th
o
d f
or
i
m
pr
ov
e
d
s
pe
e
d.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
A
ut
hor
s
s
ta
te
no f
undi
ng i
nvol
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
di
T
)
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
N
a
im
a
E
l
Y
a
nboi
y
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
M
oha
m
e
d K
ha
la
✓
✓
✓
✓
I
s
m
a
il
E
la
bba
s
s
i
✓
✓
✓
N
our
ddi
ne
E
lh
a
jr
a
t
✓
✓
O
m
a
r
E
lo
ut
a
s
s
i
✓
✓
✓
✓
✓
✓
Y
ous
s
e
f
E
l
H
a
s
s
oua
ni
✓
✓
✓
✓
C
houkr
i
M
e
s
s
a
oudi
✓
✓
✓
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
C
O
N
F
L
I
C
T
O
F
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
i
nt
e
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
,
N
.E
.Y
.,
upon r
e
a
s
ona
bl
e
r
e
que
s
t.
R
E
F
E
R
E
N
C
E
S
[
1]
F
.
C
r
e
ut
z
i
g,
P
.
A
gos
t
on,
J
.
C
.
G
ol
ds
c
hm
i
dt
,
G
.
L
ude
r
e
r
,
G
.
N
e
m
e
t
,
a
nd
R
.
C
.
P
i
e
t
z
c
ke
r
,
“
T
he
unde
r
e
s
t
i
m
a
t
e
d
pot
e
nt
i
a
l
of
s
ol
a
r
e
ne
r
gy t
o m
i
t
i
ga
t
e
c
l
i
m
a
t
e
c
ha
nge
,”
N
at
ur
e
E
ne
r
gy
, vol
. 2, no. 9, 2017, doi
:
10.
1038/
ne
ne
r
gy.2017.140.
[
2]
O
. A
. A
l
-
S
ha
hr
i
e
t
al
.
, “
S
ol
a
r
phot
ovol
t
a
i
c
e
ne
r
gy opt
i
m
i
z
a
t
i
on m
e
t
hods
, c
ha
l
l
e
nge
s
a
nd i
s
s
ue
s
:
a
c
om
pr
e
he
ns
i
v
e
r
e
vi
e
w
,”
J
ou
r
nal
of
C
l
e
ane
r
P
r
oduc
t
i
on
, vol
. 284, 2021, doi
:
10.1016/
j
.j
c
l
e
pr
o.2020.125465.
[
3]
A
.
H
.
A
l
a
m
i
e
t
al
.
,
“
M
a
na
ge
m
e
nt
of
pot
e
nt
i
a
l
c
ha
l
l
e
nge
s
of
P
V
t
e
c
hnol
ogy
pr
ol
i
f
e
r
a
t
i
on,”
Sus
t
ai
nabl
e
E
ne
r
gy
T
e
c
hnol
ogi
e
s
and
A
s
s
e
s
s
m
e
nt
s
, vol
. 51, 2022, doi
:
10.1016/
j
.s
e
t
a
.2021.101942.
[
4]
M
.
A
gha
e
i
e
t
al
.
,
“
R
e
vi
e
w
of
de
gr
a
da
t
i
on
a
nd
f
a
i
l
ur
e
phe
nom
e
na
i
n
phot
ovol
t
a
i
c
m
odul
e
s
,”
R
e
ne
w
abl
e
and
Su
s
t
ai
nabl
e
E
ne
r
g
y
R
e
v
i
e
w
s
, vol
. 159, 2022, doi
:
10.1016/
j
.r
s
e
r
.2022.112160.
[
5]
M
.
W
.
A
kr
a
m
,
G
.
L
i
,
Y
.
J
i
n,
a
nd
X
.
C
he
n,
“
F
a
i
l
ur
e
s
of
phot
ovol
t
a
i
c
m
odul
e
s
a
nd
t
he
i
r
de
t
e
c
t
i
on:
a
r
e
vi
e
w
,”
A
ppl
i
e
d
E
ne
r
gy
,
vol
.
313, 2022, doi
:
10.1016/
j
.a
pe
ne
r
gy.2022.118822.
[
6]
K
.
S
.
G
a
r
ud,
S
.
J
a
ya
r
a
j
,
a
nd
M
.
Y
.
L
e
e
,
“
A
r
e
vi
e
w
on
m
ode
l
i
ng
of
s
ol
a
r
pho
t
ovol
t
a
i
c
s
ys
t
e
m
s
us
i
ng
a
r
t
i
f
i
c
i
a
l
ne
ur
a
l
ne
t
w
or
ks
,
f
uz
z
y
l
ogi
c
,
ge
ne
t
i
c
a
l
gor
i
t
hm
a
nd
hybr
i
d
m
ode
l
s
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
E
n
e
r
gy
R
e
s
e
a
r
c
h
,
vol
.
45,
no.
1,
pp.
6
–
35,
2021,
doi
:
10.1002/
e
r
.5608.
[
7]
Y
.
Y
.
H
ong
a
nd
R
.
A
.
P
ul
a
,
“
M
e
t
hod
s
of
phot
ovol
t
a
i
c
f
a
ul
t
de
t
e
c
t
i
on
a
nd
c
l
a
s
s
i
f
i
c
a
t
i
on:
a
r
e
vi
e
w
,”
E
ne
r
gy
R
e
por
t
s
,
vol
.
8,
pp.
5898
–
5929, 2022, doi
:
10.1016/
j
.e
gyr
.2022.04.043.
[
8]
G
.
E
.
M
.
A
br
o,
A
.
A
l
i
,
S
.
A
.
M
e
m
on,
T
.
D
.
M
e
m
on,
a
nd
F
.
K
ha
n,
“
S
t
r
a
t
e
gi
e
s
a
nd
c
ha
l
l
e
nge
s
f
or
unm
a
nne
d
a
e
r
i
a
l
ve
hi
c
l
e
-
ba
s
e
d
c
ont
i
nuous
i
ns
pe
c
t
i
on
a
nd
pr
e
di
c
t
i
ve
m
a
i
nt
e
n
a
nc
e
of
s
ol
a
r
m
odul
e
s
,”
I
E
E
E
A
c
c
e
s
s
,
vol
.
12,
pp.
176615
–
176629,
2024,
doi
:
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
E
nhanc
e
d s
ol
ar
pane
ls
f
aul
t
de
te
c
ti
on appr
oac
h u
s
in
g l
ig
ht
w
e
i
ght
Y
O
L
O
…
(
N
ai
m
a E
l
Y
anboiy
)
3561
10.1109/
A
C
C
E
S
S
.2024.3505754.
[
9]
X
.
Z
ha
o,
Y
.
Z
ha
o,
S
.
H
u,
H
.
W
a
ng,
Y
.
Z
ha
ng,
a
nd
W
.
M
i
ng,
“
P
r
ogr
e
s
s
i
n
a
c
t
i
ve
i
nf
r
a
r
e
d
i
m
a
gi
ng
f
or
de
f
e
c
t
de
t
e
c
t
i
on
i
n
t
he
r
e
ne
w
a
bl
e
a
nd e
l
e
c
t
r
oni
c
i
ndus
t
r
i
e
s
,”
S
e
ns
o
r
s
, vol
. 23, no. 21, 2023, doi
:
10.33
90/
s
23218780.
[
10]
K
.
C
houdha
r
y
e
t
al
.
,
“
R
e
c
e
nt
a
dva
nc
e
s
a
nd
a
ppl
i
c
a
t
i
on
s
of
de
e
p
l
e
a
r
ni
ng
m
e
t
hods
i
n
m
a
t
e
r
i
a
l
s
s
c
i
e
nc
e
,
”
npj
C
om
put
at
i
onal
M
at
e
r
i
al
s
, vol
. 8, no. 1, 2022, doi
:
10.1038/
s
41524
-
022
-
00734
-
6.
[
11]
A
.
M
e
l
l
i
t
a
nd
S
.
K
a
l
ogi
r
ou,
“
A
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
ge
nc
e
a
nd
i
nt
e
r
ne
t
of
t
hi
ngs
t
o
i
m
pr
ove
e
f
f
i
c
a
c
y
of
di
a
gnos
i
s
a
nd
r
e
m
ot
e
s
e
ns
i
ng
of
s
ol
a
r
phot
ovol
t
a
i
c
s
y
s
t
e
m
s
:
C
ha
l
l
e
nge
s
,
r
e
c
om
m
e
nda
t
i
on
s
a
nd
f
ut
ur
e
di
r
e
c
t
i
ons
,”
R
e
ne
w
abl
e
and
Su
s
t
ai
nabl
e
E
ne
r
gy
R
e
v
i
e
w
s
,
vol
. 143, 2021, doi
:
10.1016/
j
.r
s
e
r
.2021.110889.
[
12]
S
.
Z
ha
ng,
S
.
Z
ha
ng,
B
.
W
a
ng,
a
nd
T
.
G
.
H
a
be
t
l
e
r
,
“
D
e
e
p
l
e
a
r
ni
ng
a
l
gor
i
t
hm
s
f
or
be
a
r
i
ng
f
a
ul
t
di
a
gnos
t
i
c
s
—
a
c
om
pr
e
he
ns
i
v
e
r
e
vi
e
w
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 8, pp. 29857
–
29881, 2020, doi
:
10.1109/
A
C
C
E
S
S
.2020.2972859.
[
13]
R
.
J
a
na
r
t
ha
na
n,
R
.
U
.
M
a
he
s
hw
a
r
i
,
P
.
K
.
S
hukl
a
,
P
.
K
.
S
hukl
a
,
S
.
M
i
r
j
a
l
i
l
i
,
a
n
d
M
.
K
um
a
r
,
“
I
nt
e
l
l
i
ge
nt
de
t
e
c
t
i
on
of
t
he
P
V
f
a
ul
t
s
ba
s
e
d on a
r
t
i
f
i
c
i
a
l
ne
ur
a
l
ne
t
w
or
k a
nd t
ype
2 f
uz
z
y s
ys
t
e
m
s
,”
E
ne
r
gi
e
s
, vol
. 14,
no. 20, 2021, doi
:
10.3390/
e
n14206584.
[
14]
M
.
W
.
A
kr
a
m
,
G
.
L
i
,
Y
.
J
i
n,
X
.
C
he
n,
C
.
Z
hu,
a
nd
A
.
A
hm
a
d,
“
A
ut
om
a
t
i
c
d
e
t
e
c
t
i
on
of
phot
ovol
t
a
i
c
m
odul
e
de
f
e
c
t
s
i
n
i
nf
r
a
r
e
d
i
m
a
ge
s
w
i
t
h
i
s
ol
a
t
e
d
a
nd
d
e
ve
l
op
-
m
ode
l
t
r
a
ns
f
e
r
de
e
p
l
e
a
r
ni
ng,”
Sol
a
r
E
ne
r
gy
,
vol
.
198,
pp.
175
–
186,
2020,
doi
:
10.1016/
j
.s
ol
e
ne
r
.2020.01.055.
[
15]
S
.
P
r
a
bha
ka
r
a
n,
R
.
A
.
U
t
hr
a
,
a
nd
J
.
P
r
e
e
t
ha
r
os
e
l
yn,
“
D
e
e
p
l
e
a
r
ni
ng
-
ba
s
e
d
m
ode
l
f
or
de
f
e
c
t
de
t
e
c
t
i
on
a
nd
l
oc
a
l
i
z
a
t
i
on
on
phot
ovol
t
a
i
c
pa
ne
l
s
,”
C
om
put
e
r
Sy
s
t
e
m
s
Sc
i
e
nc
e
and
E
ngi
ne
e
r
i
ng
,
v
ol
.
44,
no.
3,
pp.
2683
–
2700,
2023,
doi
:
10.32604/
c
s
s
e
.2023.028898.
[
16]
I
.
S
.
R
a
m
í
r
e
z
,
B
.
D
a
s
,
a
nd
F
.
P
.
G
.
M
á
r
que
z
,
“
F
a
ul
t
de
t
e
c
t
i
on
a
nd
di
a
gno
s
i
s
i
n
phot
ovol
t
a
i
c
pa
n
e
l
s
by
r
a
di
om
e
t
r
i
c
s
e
ns
or
s
e
m
be
dde
d i
n unm
a
nne
d a
e
r
i
a
l
ve
hi
c
l
e
s
,”
P
r
ogr
e
s
s
i
n P
hot
ov
ol
t
ai
c
s
:
R
e
s
e
a
r
c
h a
nd A
ppl
i
c
at
i
ons
, vol
. 30, no. 3, pp. 240
–
256, 2022,
doi
:
10.1002/
pi
p.3479.
[
17]
S
.
H
.
H
a
n,
T
.
R
a
hi
m
,
a
nd
S
.
Y
.
S
hi
n,
“
D
e
t
e
c
t
i
on
of
f
a
ul
t
s
i
n
s
ol
a
r
p
a
ne
l
s
u
s
i
n
g
de
e
p
l
e
a
r
ni
ng,”
i
n
2021
I
nt
e
r
nat
i
onal
C
onf
e
r
e
n
c
e
on E
l
e
c
t
r
oni
c
s
, I
nf
or
m
at
i
on, and C
om
m
uni
c
at
i
on (
I
C
E
I
C
)
, 2021, pp. 1
–
4
, doi
:
10.1109/
I
C
E
I
C
51217.2021.9369744.
[
18]
R
obof
l
ow
,
“
E
xpl
or
e
t
op
c
om
put
e
r
vi
s
i
on
m
ode
l
s
,”
R
obof
l
ow
.
2024.
A
c
c
e
s
s
e
d:
J
un.
16,
2025.
[
O
nl
i
ne
]
.
A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
r
obof
l
ow
.c
om
/
m
ode
l
s
[
19]
P
.
J
i
a
ng,
E
r
gu
D
a
j
i
,
L
i
u
F
a
ngya
o,
C
a
i
Y
i
ng,
a
nd
M
a
B
o,
“
A
r
e
vi
e
w
of
yol
o
a
l
gor
i
t
hm
de
ve
l
opm
e
nt
s
,”
P
r
oc
e
di
a
C
om
put
e
r
Sc
i
e
nc
e
, vol
. 199, pp. 1066
–
1073, 2022.
[
20]
T
.
D
i
w
a
n,
G
.
A
ni
r
udh,
a
nd
J
.
V
.
T
e
m
bhu
r
ne
,
“
O
bj
e
c
t
de
t
e
c
t
i
on
us
i
ng
Y
O
L
O
:
c
ha
l
l
e
nge
s
,
a
r
c
hi
t
e
c
t
ur
a
l
s
uc
c
e
s
s
or
s
,
da
t
a
s
e
t
s
a
n
d
a
ppl
i
c
a
t
i
ons
,”
M
ul
t
i
m
e
di
a T
ool
s
and A
ppl
i
c
at
i
ons
, vol
. 82, no. 6, pp. 9243
–
9275, 2023, doi
:
10.1007/
s
11042
-
022
-
13644
-
y.
[
21]
H
. K
. J
ung a
nd
G
. S
.
C
hoi
, “
I
m
pr
ove
d Y
O
L
O
v5:
e
f
f
i
c
i
e
nt
obj
e
c
t
de
t
e
c
t
i
on u
s
i
n
g dr
one
i
m
a
ge
s
unde
r
va
r
i
ous
c
ondi
t
i
ons
,”
A
ppl
i
e
d
Sc
i
e
nc
e
s
, vol
. 12, no. 14, 2022, doi
:
10.3390/
a
pp12147255.
[
22]
B
.
Y
a
n,
P
.
F
a
n,
X
.
L
e
i
,
Z
.
L
i
u,
a
nd
F
.
Y
a
ng,
“
A
r
e
a
l
-
t
i
m
e
a
ppl
e
t
a
r
ge
t
s
de
t
e
c
t
i
on
m
e
t
hod
f
or
pi
c
ki
ng
r
obot
ba
s
e
d
on
i
m
pr
ove
d
Y
O
L
O
v5,”
R
e
m
ot
e
Se
n
s
i
ng
, vol
. 13, no. 9, 2021, doi
:
10.3390/
r
s
13091619.
[
23]
H
.
Y
a
r
,
Z
.
A
.
K
ha
n,
F
.
U
.
M
.
U
l
l
a
h,
W
.
U
l
l
a
h,
a
nd
S
.
W
.
B
a
i
k,
“
A
m
odi
f
i
e
d
Y
O
L
O
v5
a
r
c
hi
t
e
c
t
ur
e
f
or
e
f
f
i
c
i
e
nt
f
i
r
e
de
t
e
c
t
i
on
i
n
s
m
a
r
t
c
i
t
i
e
s
,”
E
x
pe
r
t
Sy
s
t
e
m
s
w
i
t
h A
ppl
i
c
at
i
ons
, vol
. 231, 2023, doi
:
10.1016/
j
.e
s
w
a
.2023.120465.
[
24]
W
.
L
i
u,
K
.
Q
ui
j
a
no,
a
nd
M
.
M
.
C
r
a
w
f
or
d,
“
Y
O
L
O
v5
-
t
a
s
s
e
l
:
de
t
e
c
t
i
ng
t
a
s
s
e
l
s
i
n
R
G
B
U
A
V
i
m
a
ge
r
y
w
i
t
h
i
m
pr
ove
d
Y
O
L
O
v5
ba
s
e
d
on
t
r
a
ns
f
e
r
l
e
a
r
ni
ng,”
I
E
E
E
J
our
nal
of
Se
l
e
c
t
e
d
T
opi
c
s
i
n
A
ppl
i
e
d
E
ar
t
h
O
bs
e
r
v
at
i
ons
and
R
e
m
ot
e
Se
ns
i
ng
,
vol
.
15,
pp.
8085
–
8094, 2022, doi
:
10.1109/
J
S
T
A
R
S
.2022.3206399.
[
25]
G
.
N
a
i
du,
T
.
Z
uva
,
a
nd
E
.
M
.
S
i
ba
nda
,
“
A
r
e
vi
e
w
of
e
va
l
ua
t
i
on
m
e
t
r
i
c
s
i
n
m
a
c
hi
ne
l
e
a
r
ni
ng
a
l
gor
i
t
hm
s
,”
i
n
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
A
ppl
i
c
at
i
on i
n N
e
t
w
or
k
s
and Sy
s
t
e
m
s
, 2023, pp. 15
–
25
, doi
:
10.1007/
978
-
3
-
031
-
35314
-
7_2.
[
26]
P
.
H
e
nde
r
s
on
a
nd
V
.
F
e
r
r
a
r
i
,
“
E
nd
-
to
-
e
nd
t
r
a
i
ni
ng
of
obj
e
c
t
c
l
a
s
s
de
t
e
c
t
or
s
f
or
m
e
a
n
a
ve
r
a
ge
pr
e
c
i
s
i
on,”
i
n
C
om
put
e
r
V
i
s
i
on
–
A
C
C
V
2016
, 2017, pp. 198
–
213
, doi
:
10.1007/
978
-
3
-
319
-
54193
-
8_13.
B
I
O
G
R
A
P
H
I
E
S
O
F
A
U
T
H
O
R
S
Naima
El
Yanboiy
received
her
M.Sc.
degree
in
industri
al
c
omputin
g
and
instrumentation
engineer
ing from the
Faculty of
Sciences
and Tec
hniques of E
rrachidia
. She is
currently
a
Ph.D.
student
in
the
Laboratory
Optoelect
ronics
and
Appl
ied
Energy
Techniques
,
Faculty
of
Scienc
e
and
Technology
,
Moulay
Ismail
University
of
Meknes,
Errachidia,
Morocco.
Her
work
studies
and
interests
focus
Smart
fault
detection
and
prediction
in
a
photovoltaic
system
using
deep
learning
.
She
can
be
contacted
at
email:
naima.elyanboiy@
gmail.com.
Mohamed
Khala
received
graduated
with
a
master'
s
degree
in
sola
r
technologies
and
sustainable
development
from
the
Faculty
of
Scienc
es
and
Te
chniques
of
Errachidia,
Moulay
Ismail
University,
Meknes,
Morocco
in
2021.
Currently
a
P
h
.
D
.
student
at
the
same
institution
in
optoelectronics
and
applied
energy
techniques
researc
h
unit.
His
passion
for
physics
and
artificial
intelligence
(AI)
led
him
to
pursue
a
career
i
n
this
field.
He
can
b
e
contacted
at email
: khala.
mohamed@
gmail.co
m
.
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
:
3554
-
3562
3562
Ismail
Elabbassi
is
t
eacher
of
physics
and
chemistry
at
the
middl
e
school
level
received
the
master
degree
in
solar
technologies
and
sustainable
development
,
in
2021,
from
Moulay
Ismail
University,
Meknes,
Morocco,
where
he
is
currently
working
toward
the
Ph.D.
His
research
interests
include
hybrid
storage
system
modeling
a
nd
energy
management
strategies,
artificial
intell
igence
,
and
internet
of
things
.
He
can
be
contacted
at
email:
ism.elabbassi@edu.umi.ac.ma
.
Nourddin
e
Elhajrat
obtained
a
master'
s
degree
in
signals,
systems
and
computer
science
(MSSI)
from
the
Université
Sidi
Mouhamed
Ben
Abdel
ah
and
a
doctorate
in
telecommunic
ations
and
optoelectr
onics
from
the
Université
de
Moi
lay
Ismail
,
in
2014
and
2021
respectively.
He
is
currently
a
teacher
and
researcher
.
His
rese
arch
focuses
on
optical
network
design,
optical
MIMO
technology
,
and
artificial
intelligence.
He
can
be
contacted
at
email:
elhajrats
si@
gmail.co
m
.
Omar
Eloutassi
holds
a
currently
holding
the
position
of
Profes
sor
of
Highe
r
Education
at
Materials
and
Modeling
Laborato
ry
,
Department
o
f
Phys
ics
,
Faculty
of
Scienc
es
Meknes,
Moulay
Ismail
University,
Morocco
boasts
a
rich
academic
backgrou
nd
in
physics,
he
earned
his
Ph.
D
.
in
Instrumentation
and
Measurement
from
the
U
niversity
of
Bordeaux
I,
Franc
e,
in
1993,
followe
d
by
anothe
r
Doctor
ate
in
Atomic
Physica
l
Scienc
es
in
1999
from
the
Solid
State
Physics
Laboratory
at
the
Faculty
of
Sciences,
Sidi
Mohamed
Ben
Abdellah
University,
Morocco.
He
has
supervised
practical
sessions
cover
ing
a
range
of
topics
including
radio
optics,
Optoelectronics
,
diffractio
n,
and
polarization,
light
interference,
computer
architectu
re
and
vision
machine,
Metrology
and
Instru
mentation.
He
is
also
recognized
as
a
co
-
author
of
several
national
and
international
p
ublications.
He
can
b
e
contacted
at
email
:
eloutass
iomar@
gmail.co
m.
Youssef
El
Hassouan
i
is
an
academic
researcher
at
the
Departme
nt
of
Physics,
Faculty
of
Scienc
es
and
Techn
iques,
Moulay
Ismail
Univer
sity,
E
rrachidia
,
Morocco
.
He
holds
a
doctoral
degree
from
the
University
of
Lille
1
(France)
and
the
University
Mohammed
Premie
r
Oujda
(Moroc
co).
His
resea
rch
focuse
s
on
phononic
and
p
hotonic
crysta
ls,
with
a
particular
interest
in
energy
photovoltaics.
He
has
supervised
several
doctoral
theses
and
co
-
authored
numerous
national
and
internati
onal
publicat
ions.
He
can
be
contacted
at
email:
hassouani@
yahoo.fr.
Choukri
Messaoudi
is
a
Professor
of
Higher
Education
at
t
he
Faculty
of
Scienc
es
and
Techn
iques
in
Erra
chidia
,
Morocc
o,
specia
lizing
in
re
newable
energies
.
With
expertise
in
materials
physics
and
renewable
energies,
he
serves
as
b
oth
a
researcher
and
the
coordinat
or
of
the
master'
s
program
in
solar
technology
and
sustai
nable
development
.
His
research
interests
span
various
topics
including
concentrated
sola
r
power,
heat
transfer,
photovoltaic systems, and solar power
.
He can be contacted at email:
messaoudic2
@
yahoo.fr.
Evaluation Warning : The document was created with Spire.PDF for Python.