I
nte
rna
t
io
na
l J
o
urna
l o
f
Adv
a
nces in Applie
d Science
s
(
I
J
AAS)
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
2017
,
p
p
.
3
0
3
~3
1
2
I
SS
N:
2252
-
8814
3
03
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
jo
u
r
n
a
l.c
o
m/o
n
lin
e/in
d
ex
.
p
h
p
/I
J
AAS
Rev
iew
of Ma
chi
ne Visio
n
Ba
sed
I
nsula
tor Ins
pecti
o
n Sys
te
m
s
for O
v
er
hea
d P
o
w
er Distr
ibu
tion
Sy
ste
m
P
.
Su
ry
a
P
ra
s
a
d
1
,
B
.
P
ra
bh
a
k
a
ra
Ra
o
2
1
De
p
a
rtme
n
t
o
f
ECE
,
M
V
G
R
Co
l
leg
e
o
f
En
g
in
e
e
rin
g
,
V
izia
n
a
g
a
ra
m
,
A
n
d
h
ra
P
ra
d
e
sh
,
In
d
ia
2
De
p
a
rtme
n
t
o
f
ECE
,
JN
T
U Ka
k
i
n
a
d
a
,
A
n
d
h
ra
P
ra
d
e
sh
,
I
n
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Sep
1
3
,
2
0
1
7
R
ev
i
s
ed
No
v
1
3
,
2
0
1
7
A
cc
ep
ted
No
v
2
1
,
2
0
1
7
T
h
e
n
e
c
e
ss
it
y
to
h
a
v
e
re
li
a
b
le
a
n
d
q
u
a
li
ty
p
o
w
e
r
d
istri
b
u
ti
o
n
is
in
c
re
a
sin
g
,
a
n
d
h
e
n
c
e
th
e
re
is
g
re
a
t
sc
o
p
e
f
o
r
re
se
a
rc
h
o
n
a
u
t
o
m
a
ti
o
n
o
f
d
istri
b
u
t
io
n
s
y
ste
m
.
T
h
e
re
a
re
sig
n
s
o
f
in
c
r
e
a
se
d
re
se
a
r
c
h
in
t
h
e
w
o
rk
o
n
c
o
n
d
it
io
n
m
o
n
it
o
rin
g
o
f
in
s
u
lato
rs
d
u
r
in
g
t
h
e
la
st
f
e
w
d
e
c
a
d
e
s.
T
h
e
p
o
ss
ib
l
e
f
a
il
u
re
s
c
a
n
b
e
p
re
d
icte
d
b
e
f
o
re
th
e
y
a
c
tu
a
ll
y
o
c
c
u
r
b
y
u
sin
g
th
e
c
o
n
d
it
i
o
n
m
o
n
it
o
rin
g
o
f
c
a
b
les
o
r
a
n
y
e
le
c
tri
c
a
l
e
q
u
ip
m
e
n
t
o
n
-
li
n
e
.
T
h
o
se
a
ss
e
ts
su
c
h
a
s
to
we
rs,
c
o
n
d
u
c
to
rs
a
n
d
in
s
u
lato
rs
w
h
ich
a
re
o
n
th
e
th
re
sh
o
ld
o
f
fa
il
u
re
h
a
v
e
to
b
e
re
p
lac
e
d
o
r
re
p
a
ired
,
so
t
h
a
t
f
o
rc
e
d
o
u
tag
e
s
re
d
u
c
e
.
T
ra
d
it
io
n
a
ll
y
th
e
w
o
rk
e
rs
w
h
o
in
sp
e
c
t
th
e
se
li
n
e
s
c
h
e
c
k
th
e
m
in
c
lo
se
p
ro
x
im
it
y
b
y
g
o
in
g
f
o
r
f
o
o
t
-
p
a
tro
ll
i
n
g
a
n
d
p
o
le
-
c
li
m
b
in
g
.
W
it
h
a
n
in
c
re
d
ib
le
e
x
p
a
n
sio
n
o
f
p
o
w
e
r
d
istri
b
u
ti
o
n
n
e
tw
o
rk
e
v
e
n
to
re
m
o
te
a
re
a
s,
p
re
v
io
u
sly
m
e
n
ti
o
n
e
d
m
e
th
o
d
s
d
o
n
o
t
se
e
m
to
b
e
v
iab
le.
In
d
e
v
e
lo
p
e
d
c
o
u
n
tri
e
s
a
e
rial
p
a
tro
l
li
n
g
h
a
s
b
e
e
n
a
d
o
p
te
d
to
m
o
n
it
o
r
t
h
e
in
s
u
lato
r
s
a
s
a
n
a
lt
e
rn
a
ti
v
e
.
T
h
e
d
e
v
e
lo
p
m
e
n
t
o
f
a
n
e
ff
ici
e
n
t
m
e
th
o
d
o
f
c
o
n
d
it
io
n
m
o
n
it
o
ri
n
g
b
y
u
sin
g
im
a
g
e
p
ro
c
e
ss
in
g
f
o
ll
o
w
e
d
b
y
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
iq
u
e
s
is
f
o
u
n
d
to
b
e
a
su
it
a
b
le
m
e
th
o
d
a
n
d
th
u
s
e
m
e
rg
in
g
a
s
a
fe
a
sib
le
o
p
ti
o
n
f
o
r
re
a
l
-
ti
m
e
i
m
p
le
m
e
n
tatio
n
.
T
h
is
re
v
ie
w
p
a
p
e
r
c
o
v
e
rs
o
v
e
ra
ll
a
sp
e
c
ts
o
f
a
u
to
m
a
ti
c
d
e
tec
ti
o
n
o
f
d
e
fe
c
ts
o
f
in
su
lato
r
s
y
ste
m
s
o
f
e
le
c
tri
c
p
o
w
e
r
li
n
e
s
a
n
d
c
las
sif
ica
ti
o
n
in
to
d
if
f
e
r
e
n
t
c
las
se
s
b
y
u
sin
g
v
isio
n
-
b
a
se
d
tec
h
n
iq
u
e
s.
K
ey
w
o
r
d
:
C
las
s
i
f
icatio
n
C
o
n
d
itio
n
m
o
n
ito
r
in
g
Def
ec
t d
etec
tio
n
Featu
r
e
ex
tr
ac
tio
n
I
m
ag
e
p
r
o
ce
s
s
i
n
g
Ma
ch
i
n
e
lear
n
i
n
g
Co
p
y
rig
h
t
©
201
7
In
s
t
it
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
P
.
S
u
ry
a
P
ra
sa
d
,
De
p
a
rtme
n
t
o
f
ECE
,
M
V
G
R
Co
ll
e
g
e
o
f
En
g
in
e
e
rin
g
,
V
izia
n
a
g
a
ra
m
,
A
n
d
h
ra
P
ra
d
e
sh
,
I
n
d
ia
.
Em
a
il
:
su
r
y
a
p
ra
sa
d
p
@
y
a
h
o
o
.
c
o
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
r
o
u
g
h
o
u
t
w
o
r
ld
w
id
e
elec
t
r
ic
p
o
w
er
u
t
ilit
ie
s
ar
e
ad
o
p
tin
g
th
e
co
m
p
u
ter
aid
ed
m
o
n
ito
r
in
g
,
m
an
a
g
e
m
e
n
t
an
d
co
n
tr
o
l
o
f
elec
tr
ic
p
o
w
er
d
is
tr
ib
u
t
io
n
s
y
s
t
e
m
m
o
r
e
an
d
m
o
r
e,
to
p
r
o
v
id
e
b
etter
s
er
v
ices
to
th
e
co
n
s
u
m
er
s
o
f
elec
tr
icit
y
[
1
]
.
T
h
e
id
ea
o
f
d
is
tr
ib
u
tio
n
a
u
to
m
at
io
n
b
eg
a
n
in
1
9
7
0
s
an
d
t
h
e
m
o
tiv
at
io
n
to
i
m
p
r
o
v
e
o
p
er
atin
g
p
er
f
o
r
m
a
n
ce
o
f
d
is
tr
ib
u
tio
n
s
y
s
te
m
s
w
a
s
d
er
iv
ed
f
r
o
m
th
e
u
s
a
g
e
o
f
co
m
p
u
ter
an
d
co
m
m
u
n
icatio
n
s
tech
n
o
lo
g
y
.
T
h
en
o
n
w
ar
d
s
,
th
e
i
m
p
r
o
v
e
m
en
t
i
n
au
to
m
atio
n
o
f
d
is
tr
ib
u
ti
o
n
s
e
s
t
y
m
h
a
s
b
ee
n
d
ictated
b
y
i
n
cr
ea
s
i
n
g
co
m
p
le
x
it
y
lev
e
l
s
i
n
t
h
e
ex
i
s
ti
n
g
tech
n
o
lo
g
ies
f
o
r
m
o
n
ito
r
in
g
,
co
n
tr
o
l,
co
m
m
u
n
ica
tio
n
tech
n
o
lo
g
ies
a
n
d
also
th
e
p
er
f
o
r
m
an
ce
a
n
d
co
s
t
o
f
th
e
e
q
u
ip
m
e
n
t
[
2
4
]
.
T
o
en
s
u
r
e
u
n
in
ter
r
u
p
ted
r
eliab
le
o
p
er
atio
n
o
f
p
o
w
er
d
is
tr
ib
u
t
io
n
s
y
s
te
m
,
th
e
p
er
f
o
r
m
a
n
c
e
o
f
in
s
u
lato
r
s
u
n
d
er
d
if
f
er
e
n
t
en
v
ir
o
n
m
e
n
tal
co
n
d
itio
n
s
m
u
s
t
b
e
r
eliab
le
an
d
s
atis
f
ac
to
r
y
.
So
,
th
e
s
u
p
p
lier
h
as
to
ta
k
e
all
t
h
e
n
ec
e
s
s
ar
y
s
tep
s
to
in
s
p
ec
t
h
i
s
in
s
ta
llatio
n
s
an
d
e
n
s
u
r
e
th
at
t
h
eir
p
er
f
o
r
m
a
n
ce
co
m
p
l
y
w
it
h
th
e
r
e
g
u
latio
n
s
.
T
h
e
o
th
er
r
ea
s
o
n
s
to
p
er
f
o
r
m
in
s
p
ec
tio
n
s
o
f
l
in
e
s
r
eg
u
lar
l
y
ar
e:
i.
Dete
ctio
n
o
f
li
n
e
d
e
f
ec
t
s
ea
r
l
y
r
ed
u
ce
s
p
o
w
er
c
u
t
s
a
n
d
it
p
av
e
s
t
h
e
w
a
y
f
o
r
g
o
o
d
cu
s
to
m
er
ca
r
e
a
n
d
a
d
v
an
ta
g
e
o
v
er
o
t
h
er
co
m
p
etit
o
r
s
;
ii.
Hav
in
g
a
p
h
o
to
g
r
ap
h
i
c
r
ec
o
r
d
allo
w
s
t
h
e
r
ep
air
team
s
to
k
n
o
w
ex
ac
t
l
y
ab
o
u
t
th
e
r
ep
air
an
d
m
ai
n
te
n
a
n
ce
to
b
e
d
o
n
e
w
h
ic
h
r
ed
u
ce
s
o
p
er
atio
n
al
co
s
ts
;
iii.
P
u
b
lic
ar
e
p
r
o
tecte
d
b
y
d
etec
tin
g
th
e
m
is
s
in
g
o
r
b
r
o
k
en
s
af
et
y
f
ea
t
u
r
es
o
n
p
o
les
[
2
]
.
I
f
a
d
ev
ice
is
f
o
u
n
d
to
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
201
7
:
3
0
3
–
3
1
2
304
b
e
th
e
ca
u
s
e
o
f
th
e
f
a
u
lt,
r
ep
air
o
r
r
ep
lace
m
en
t
o
f
th
e
s
a
m
e
co
u
ld
b
e
m
ad
e
o
r
r
ec
o
m
m
e
n
d
atio
n
s
ar
e
g
iv
e
n
f
o
r
an
y
o
t
h
er
r
eq
u
ir
ed
m
ain
ten
a
n
c
e
ac
tio
n
s
[
3
]
.
I
f
t
h
e
d
is
tr
ib
u
tio
n
eq
u
ip
m
en
t
s
tar
ts
w
ea
k
e
n
i
n
g
,
it
ca
n
b
e
a
n
ticip
ated
t
h
at
a
n
u
n
p
r
ed
ictab
le
f
a
u
lt
is
d
ev
elo
p
in
g
i
n
t
h
e
s
y
s
te
m
m
a
y
b
e
f
r
o
m
s
e
v
er
al
d
a
y
s
o
r
s
ev
e
r
al
m
o
n
th
s
.
So
,
i
n
o
r
d
er
to
h
a
v
e
a
r
elia
b
le
p
o
w
e
r
d
is
tr
ib
u
tio
n
s
y
s
te
m
,
it
i
s
cr
u
c
ial
to
m
o
n
ito
r
d
i
s
tr
ib
u
tio
n
,
cl
ass
i
f
y
a
n
d
id
en
ti
f
y
t
h
e
v
ar
io
u
s
t
y
p
e
s
o
f
f
ail
u
r
es
b
ef
o
r
e
th
e
ac
tu
al
b
r
ea
k
d
o
w
n
o
r
b
r
ea
k
o
u
t
o
f
th
e
eq
u
ip
m
en
t
o
cc
u
r
s
[
3
]
.
T
h
e
o
n
-
l
in
e
p
r
o
ce
s
s
in
g
to
m
o
n
ito
r
t
h
e
h
ea
lt
h
o
f
eq
u
ip
m
e
n
t
is
g
e
n
e
r
all
y
ca
lled
as
co
n
d
itio
n
m
o
n
i
to
r
in
g
(
C
M)
[
4
]
.
T
h
e
o
v
er
h
e
ad
lin
es
h
av
e
m
a
n
y
ite
m
s
s
u
c
h
as
i
n
s
u
lato
r
s
,
co
n
d
u
cto
r
s
an
d
fi
tti
n
g
s
,
w
h
ic
h
ca
n
b
e
co
n
tin
u
o
u
s
l
y
m
o
n
i
to
r
ed
o
n
lin
e
[
5
–
1
1
]
.
So
,
th
er
e
s
h
o
u
ld
b
e
a
co
n
s
tan
t
m
o
n
ito
r
in
g
o
f
th
e
s
y
s
te
m
to
tr
a
ce
its
p
r
o
g
r
ess
io
n
f
r
o
m
h
ea
lt
h
y
to
s
ic
k
s
tat
u
s
.
A
m
en
tio
n
ab
o
u
t
h
ea
lt
h
i
n
d
ex
b
a
s
ed
o
n
m
o
n
ito
r
in
g
d
ata
is
p
r
esen
ted
i
n
[
1
2
]
.
So
,
t
h
e
d
ev
elo
p
m
en
t
o
f
tr
a
n
s
d
u
ce
r
tech
n
o
lo
g
ies,
co
m
p
u
ter
tec
h
n
o
lo
g
ies
a
n
d
s
i
g
n
a
l
p
r
o
ce
s
s
i
n
g
tech
n
iq
u
es
alo
n
g
w
i
th
ar
ti
f
ic
i
al
–
in
telli
g
e
n
ce
(
A
I
)
tech
n
i
q
u
es
h
as
m
ad
e
it
p
o
s
s
ib
le
to
im
p
le
m
e
n
t
C
M
m
o
r
e
e
f
f
e
ctiv
el
y
o
n
elec
tr
ical
eq
u
ip
m
en
t
[
1
3
]
.
T
h
is
r
ev
ie
w
ai
m
s
at
co
m
p
ar
i
n
g
a
n
d
co
n
tr
as
tin
g
th
e
r
ep
o
r
ted
r
esear
ch
th
at
w
a
s
d
o
n
e
b
ased
o
n
I
m
a
g
e
P
r
o
ce
s
s
i
n
g
alo
n
g
w
it
h
Ma
ch
i
n
e
L
ea
r
n
i
n
g
T
ec
h
n
iq
u
es
in
co
m
p
u
tin
g
s
o
m
e
m
e
an
in
g
f
u
l
i
n
f
o
r
m
atio
n
o
b
tai
n
ed
f
r
o
m
co
n
d
itio
n
m
o
n
ito
r
i
n
g
.
2.
CO
M
P
L
E
XIT
I
E
S O
F
AU
T
O
M
AT
I
O
N
O
F
I
NSU
L
A
T
O
R
M
O
NIT
O
RING
A
cc
u
r
ate
a
n
d
r
eliab
le
tech
n
iq
u
es
to
lo
ca
te,
d
etec
t
f
ai
lu
r
es,
i
d
en
tific
atio
n
a
n
d
e
v
alu
a
tio
n
o
f
s
e
v
er
it
y
,
h
av
e
b
ec
o
m
e
cr
u
cia
l
to
i
n
d
ea
lin
g
w
it
h
t
h
e
m
ai
n
te
n
an
ce
wo
r
k
in
t
h
e
s
ch
ed
u
led
ti
m
e
a
n
d
r
ea
lis
tic
co
s
t
[
4
]
.
On
-
s
ite
Ma
n
u
a
l
d
etec
tio
n
is
co
s
tl
y
,
ti
m
e
co
n
s
u
m
in
g
,
an
d
is
an
i
m
p
r
ac
tical
tas
k
in
ca
s
e
o
f
m
o
n
ito
r
i
n
g
t
h
e
lo
n
g
lin
es
s
p
r
ea
d
o
v
er
a
lar
g
e
d
is
ta
n
ce
an
d
d
i
f
f
icu
lt
ter
r
ain
s
.
Vid
eo
s
u
r
v
ei
llan
ce
s
y
s
te
m
b
ased
o
n
i
m
a
g
e
p
r
o
ce
s
s
i
n
g
ca
n
d
o
th
at
t
y
p
e
o
f
m
o
n
i
to
r
in
g
v
er
y
ea
s
il
y
.
U
s
e
o
f
h
el
ico
p
ter
s
h
a
s
b
ee
n
p
r
o
p
o
s
ed
f
o
r
v
id
eo
s
u
r
v
eilla
n
ce
i
n
UK
to
in
s
p
ec
t
t
h
e
d
is
tr
ib
u
tio
n
s
y
s
te
m
.
J
o
n
es
a
n
d
E
ar
p
[
1
5
]
g
av
e
d
etailed
d
is
cu
s
s
io
n
o
n
t
h
e
m
o
tiv
a
tio
n
to
g
o
f
o
r
v
id
eo
in
s
p
ec
tio
n
tec
h
n
iq
u
es
a
n
d
th
e
p
r
o
b
lem
s
th
at
m
a
y
ar
is
e.
A
n
d
th
er
e
is
a
m
e
n
tio
n
o
f
w
id
e
ap
p
licatio
n
o
f
ae
r
ial
in
s
p
ec
tio
n
[
1
6
]
f
o
r
co
n
d
itio
n
m
o
n
ito
r
in
g
o
f
o
v
er
h
ea
d
l
in
es.
Au
to
m
a
tic
v
id
eo
s
u
r
v
e
illa
n
ce
u
s
in
g
a
h
elico
p
ter
o
f
p
o
w
er
lin
es
is
n
o
t
as
s
tr
ai
g
h
t
f
o
r
w
a
r
d
an
d
th
e
p
r
o
b
lem
s
ar
e:
a.
P
atter
n
r
ec
o
g
n
itio
n
ap
p
li
ed
to
tar
g
et
lo
ca
tio
n
s
[
1
7
]
,
b
.
Stab
ilizatio
n
th
e
ca
m
er
a
i
n
co
m
p
e
n
s
at
io
n
o
f
t
h
e
h
elico
p
te
r
’
s
6
d
e
g
r
ee
-
o
f
-
f
r
ee
d
o
m
(
DO
F
)
m
o
v
e
m
e
n
t
[
1
8
,
1
9
]
,
c.
A
cq
u
ir
e
an
d
m
ai
n
tai
n
t
h
e
tar
g
et
in
th
e
ca
m
er
a’
s
fi
eld
o
f
v
ie
w
(
FOV)
[
1
7
]
,
iv
)
C
a
m
er
a’
s
r
esid
u
al
s
i
g
h
tli
n
e
m
o
tio
n
r
esu
lt
s
i
m
a
g
e
d
eg
r
ad
atio
n
[
1
8
,
1
5
,
2
0
]
,
an
d
v
)
Data
an
al
y
s
is
s
y
s
te
m
.
T
h
e
in
s
p
ec
tio
n
p
r
o
ce
s
s
b
ec
o
m
es
au
to
m
atic
b
y
t
h
e
u
s
e
o
f
v
id
eo
s
u
r
v
e
illa
n
ce
tech
n
iq
u
e
s
an
d
it
also
i
m
p
r
o
v
es
it
s
d
ep
th
an
d
co
v
er
ag
e
b
ec
a
u
s
e
it
p
r
o
v
id
es
p
er
m
a
n
e
n
t
r
ec
o
r
d
o
f
th
e
i
m
a
g
es
[
1
4
]
.
B
u
t,
t
h
e
i
n
s
p
ec
tio
n
o
f
i
n
s
u
lato
r
s
in
r
ea
l
-
ti
m
e
f
ac
es
a
n
u
m
b
er
o
f
c
h
alle
n
g
e
s
as
f
o
llo
w
s
:
I
m
a
g
e
b
lu
r
r
i
n
g
,
ca
m
e
r
a
s
ig
h
t
co
n
tr
o
l,
f
ast
ch
a
n
g
i
n
g
b
ac
k
g
r
o
u
n
d
,
an
d
g
r
ad
u
a
l
in
tr
u
s
io
n
o
f
tr
ee
li
m
b
s
in
to
th
e
o
v
er
h
ea
d
p
o
w
er
lin
e
s
[
9
]
.
A
s
an
i
m
p
r
o
v
e
m
e
n
t,
v
id
eo
s
u
r
v
eilla
n
ce
w
it
h
f
ix
ed
ca
m
er
as
as
ap
p
lied
f
o
r
p
ed
estrian
d
etec
tio
n
[
2
1
]
an
d
th
is
s
ee
m
s
to
b
e
a
p
r
o
m
i
s
in
g
s
o
lu
t
io
n
f
o
r
v
id
eo
s
u
r
v
eillan
ce
o
f
p
o
w
er
d
is
tr
ib
u
tio
n
s
y
s
te
m
in
s
u
lato
r
s
.
W
ith
th
e
f
a
s
t
c
h
a
n
g
i
n
g
s
ce
n
a
r
io
in
a
v
ailab
ilit
y
o
f
c
h
ea
p
er
d
ig
ital
ca
m
er
a
s
w
it
h
g
o
o
d
p
er
f
o
r
m
a
n
ce
,
t
h
e
m
o
u
n
ti
n
g
o
f
v
id
eo
ca
m
er
as
b
ec
a
m
e
c
h
ea
p
.
B
u
t
th
e
m
an
p
o
w
er
b
ec
o
m
e
s
v
er
y
ex
p
en
s
iv
e
to
p
er
s
o
n
all
y
o
b
s
er
v
e
an
d
i
n
ter
p
r
et
t
h
e
r
esu
lts
.
S
o
it
w
a
s
p
r
o
p
o
s
ed
to
o
p
er
ate
ca
m
er
as
at
r
e
g
u
lar
in
ter
v
a
ls
o
f
ti
m
e
to
g
et
i
m
ag
es
o
f
p
o
w
er
d
is
tr
ib
u
tio
n
lin
e
s
alo
n
g
w
i
th
i
n
s
u
lato
r
s
,
s
e
n
t
f
o
r
an
al
y
s
is
to
t
h
e
co
n
tr
o
l r
o
o
m
b
y
t
h
e
u
s
e
o
f
r
em
o
te
ter
m
i
n
al
u
n
i
t (
R
T
U)
s
[
9
]
an
d
th
is
m
et
h
o
d
ass
u
r
es p
r
o
m
is
in
g
r
es
u
lt
s
.
3.
P
RIOR
L
I
T
E
RA
T
URE R
E
VIE
W
I
n
t
h
e
la
s
t
f
e
w
d
ec
ad
es,
t
h
er
e
h
a
s
b
ee
n
a
lo
t
o
f
r
esear
ch
co
n
d
u
cted
i
n
th
e
fi
eld
o
f
au
to
m
atic
p
o
w
er
lin
e
i
n
s
p
ec
tio
n
.
A
r
ev
ie
w
p
ap
er
,
ad
d
r
ess
in
g
e
x
cl
u
s
i
v
el
y
t
h
e
in
s
p
ec
tio
n
w
it
h
ae
r
ial
v
eh
ic
les
is
g
iv
e
n
i
n
[
2
5
]
.
A
s
u
r
v
e
y
p
ap
er
h
as
b
ee
n
p
u
b
lis
h
ed
ab
o
u
t
th
e
o
v
e
r
h
ea
d
p
o
w
er
lin
e
i
n
s
p
ec
tio
n
w
h
ic
h
in
clu
d
e
s
au
to
m
a
ted
h
elico
p
ter
-
as
s
is
ted
in
s
p
ec
tio
n
b
ased
o
n
in
s
p
ec
tio
n
w
i
th
fly
in
g
a
n
d
cli
m
b
i
n
g
r
o
b
o
ts
[
2
7
]
.
O
v
er
th
e
y
ea
r
s
,
th
er
e
ar
e
a
n
u
m
b
er
o
f
r
e
v
ie
w
p
ap
er
s
o
n
co
n
d
itio
n
m
o
n
it
o
r
in
g
o
f
d
is
tr
ib
u
tio
n
p
o
w
e
r
lin
e
in
s
p
ec
tio
n
[2
-
4
,
2
3
-
2
7
,
3
2
,
3
8
]
.
A
p
r
esen
ta
tio
n
o
n
f
ea
s
ib
ilit
y
s
tu
d
y
ab
o
u
t
c
h
ar
ac
ter
izatio
n
o
f
e
m
er
g
i
n
g
in
s
u
lato
r
f
ail
u
r
e
to
p
r
e
d
ict
f
au
lt
i
n
t
h
e
d
is
tr
ib
u
tio
n
i
s
g
iv
en
i
n
[
2
]
.
A
b
r
ie
f
o
u
tli
n
e
o
n
o
v
er
h
ea
d
li
n
e
d
eter
io
r
atio
n
,
av
ailab
le
i
n
s
p
ec
tio
n
m
et
h
o
d
s
an
d
in
f
o
r
m
atio
n
ab
o
u
t
a
p
r
o
j
ec
t
b
ein
g
u
n
d
er
ta
k
en
b
y
th
e
P
o
w
er
an
d
E
n
er
g
y
S
y
s
te
m
s
R
esear
c
h
Gr
o
u
p
at
t
h
e
U
n
i
v
er
s
it
y
o
f
B
ath
w
h
o
m
o
n
ito
r
ed
o
v
er
h
ea
d
li
n
es
o
n
-
lin
e
[
4
]
.
T
h
e
d
ev
elo
p
m
en
t
o
f
a
n
e
w
t
w
o
-
co
u
r
s
e
s
eq
u
e
n
ce
to
r
e
f
lect
t
h
e
r
ad
ical
ch
an
g
es
t
h
at
o
cc
u
r
s
o
r
ex
p
ec
ted
to
h
ap
p
en
i
n
f
u
t
u
r
e
w
a
s
r
ec
o
m
m
en
d
ed
b
y
th
e
au
th
o
r
i
n
[
2
3
]
.
An
a
u
to
m
atic
v
id
eo
s
u
r
v
e
illa
n
ce
s
y
s
te
m
u
s
i
n
g
a
m
a
n
n
ed
h
elico
p
ter
w
a
s
p
r
o
p
o
s
ed
to
b
e
a
p
r
o
m
is
i
n
g
alter
n
ati
v
e
f
o
r
tr
ad
itio
n
al
in
s
p
ec
tio
n
m
et
h
o
d
s
o
f
p
o
w
er
li
n
es
i
n
[
2
5
]
an
d
r
e
m
o
tel
y
o
p
er
ated
u
n
m
a
n
n
ed
fly
i
n
g
r
o
b
o
t
w
as
a
n
ticip
ated
as
t
h
e
f
u
tu
r
e
o
f
o
v
er
h
ea
d
p
o
w
er
lin
e
in
s
p
ec
tio
n
.
B
.
A
v
id
ar
i
n
[
2
6
]
g
iv
e
s
b
r
ief
e
x
p
la
n
atio
n
o
f
t
h
e
in
s
p
ec
tio
n
m
eth
o
d
s
i
n
w
o
r
k
s
an
d
f
o
c
u
s
es
o
n
t
h
e
elec
tr
o
n
ic
ap
p
r
o
ac
h
.
He
also
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
R
ev
iew
o
f Ma
ch
in
e
V
is
io
n
b
a
s
ed
I
n
s
u
la
to
r
I
n
s
p
ec
tio
n
S
yste
ms fo
r
Ove
r
h
ea
d
.
..
(
P
.
S
u
r
ya
P
r
a
s
a
d
)
305
d
is
cu
s
s
es
t
h
e
co
n
ce
p
t
o
f
a
a
ir
b
o
r
n
e,
co
m
p
letel
y
s
ta
n
d
a
n
d
ele
ctr
o
n
ic
m
et
h
o
d
[
2
6
]
.
T
h
e
m
o
s
t
p
r
o
m
i
n
en
t
ac
h
iev
e
m
e
n
t
s
ab
o
u
t
in
s
p
ec
ti
o
n
o
f
p
o
w
er
d
is
tr
ib
u
t
io
n
lin
e
b
y
m
o
b
ile
r
o
b
o
ts
ar
e
p
r
es
en
ted
in
[
2
7
]
.
T
h
e
in
s
u
f
f
icien
c
y
o
f
tr
ad
itio
n
al
wa
y
s
o
f
i
n
s
u
lato
r
d
etec
tio
n
h
a
s
h
ea
d
ed
to
w
ar
d
s
lo
t
o
f
r
esear
ch
o
n
a
u
to
m
atic
o
n
-
lin
e
d
e
tectio
n
m
eth
o
d
.
Am
o
n
g
t
h
e
v
ar
io
u
s
m
et
h
o
d
s
o
f
d
etec
tio
n
s
u
r
v
e
y
ed
in
[
3
8
]
,
th
e
elec
tr
ic
f
ield
d
is
tr
ib
u
tio
n
m
et
h
o
d
d
etec
ts
t
h
e
in
ter
n
al
i
n
s
u
latio
n
d
ef
ec
t
s
li
v
e
li
n
e.
An
ex
ce
llen
t
th
eo
r
eti
ca
l
b
ac
k
g
r
o
u
n
d
to
i
m
a
g
e
p
r
o
ce
s
s
in
g
is
co
v
er
ed
b
y
Go
n
za
lez
an
d
W
o
o
d
s
[
3
0
]
an
d
class
if
ica
tio
n
b
ased
o
n
n
eu
r
al
n
et
w
o
r
k
is
g
iv
e
n
Ha
y
k
in
s
[
3
1
]
.
Av
ailab
ilit
y
o
f
t
h
e
p
u
b
lis
h
ed
liter
atu
r
e
o
n
a
u
to
m
ated
m
o
n
ito
r
in
g
o
f
i
n
s
u
lato
r
s
’
co
n
d
it
io
n
m
ain
l
y
co
n
s
is
ts
o
f
r
esear
ch
w
o
r
k
d
o
n
e
at
s
e
v
er
al
ac
ad
e
m
ic
an
d
r
esear
ch
i
n
s
ti
tu
t
io
n
s
.
A
ca
d
e
m
ic
i
n
s
t
itu
tes
h
a
v
e
p
u
b
lis
h
ed
n
u
m
b
er
o
f
r
esear
c
h
w
o
r
k
s
.
T
h
er
e
ar
e
s
o
m
e
p
a
p
er
s
p
u
b
lis
h
ed
i
n
r
ep
o
r
ted
r
e
s
ea
r
ch
w
o
r
k
m
ai
n
l
y
f
o
cu
s
es
o
n
clas
s
i
f
icatio
n
o
f
d
ef
ec
ts
o
f
i
n
s
u
la
to
r
s
o
f
p
o
w
er
d
is
tr
ib
u
tio
n
,
i
m
p
le
m
e
n
ted
u
s
i
n
g
v
ar
io
u
s
k
i
n
d
s
o
f
ap
p
r
o
ac
h
es.
Ho
w
ev
er
,
t
h
e
au
t
h
o
r
s
c
o
u
ld
n
o
t
f
i
n
d
an
y
r
ev
ie
w
p
ap
er
o
n
th
e
r
e
s
ea
r
ch
w
o
r
k
d
o
n
e
in
t
h
e
f
ield
o
f
co
n
d
itio
n
m
o
n
ito
r
in
g
a
n
d
clas
s
if
ica
tio
n
o
f
i
n
s
u
la
to
r
s
o
f
d
is
tr
ib
u
tio
n
s
y
s
te
m
u
s
in
g
i
m
a
g
e
p
r
o
ce
s
s
in
g
co
m
b
in
ed
w
it
h
ar
ti
f
icial
m
ac
h
in
e
lear
n
i
n
g
tech
n
iq
u
es.
T
h
er
ef
o
r
e,
th
is
a
tte
m
p
t
is
b
e
i
n
g
d
o
n
e
to
co
n
s
o
lid
ate
th
e
p
u
b
lis
h
ed
liter
atu
r
e
f
r
o
m
ac
ad
e
m
ia
a
n
d
r
esear
ch
in
s
tit
u
tes
o
n
t
h
e
to
p
ic
o
f
au
to
m
at
ic
in
s
p
ec
tio
n
o
f
th
e
co
n
d
itio
n
o
f
in
s
u
lato
r
s
.
I
t
t
h
en
t
h
r
o
w
s
a
li
g
h
t
o
n
v
ar
io
u
s
d
etec
tio
n
a
n
d
a
n
al
y
s
i
s
tec
h
n
iq
u
es
p
r
esen
tl
y
b
ein
g
u
s
ed
.
T
h
e
ai
m
of
th
i
s
p
ap
er
is
to
co
m
p
ar
e
a
n
d
co
n
tr
ast
t
h
e
an
a
l
y
s
es
ai
m
e
d
to
d
etec
t
an
d
class
if
y
t
h
e
d
ef
ec
ted
o
r
cr
ac
k
ed
in
s
u
lato
r
s
an
d
th
u
s
co
n
tr
ib
u
te
to
t
h
e
d
esi
g
n
a
n
d
d
ep
lo
y
m
en
t
o
f
an
o
n
-
li
n
e
co
n
d
itio
n
m
o
n
ito
r
in
g
t
h
r
o
u
g
h
th
e
u
s
e
o
f
m
ac
h
in
e
v
i
s
io
n
tec
h
n
iq
u
es [
4
]
.
4.
CAT
E
G
O
R
I
E
S O
F
I
NSU
L
AT
O
RS A
N
D
DE
F
E
CT
S
T
h
e
in
s
u
la
to
r
s
u
s
ed
i
n
tr
an
s
m
is
s
io
n
li
n
e
ar
e
th
e
d
ev
ices
w
h
ich
ar
e
u
s
ed
to
co
n
tain
,
s
u
p
p
o
r
t
o
r
s
ep
ar
ate
th
e
elec
tr
ical
co
n
d
u
cto
r
s
.
T
h
e
y
ar
e
u
s
ed
f
o
r
h
ig
h
v
o
lta
g
e
p
o
w
er
d
is
tr
ib
u
tio
n
n
et
w
o
r
k
s
.
T
h
e
tr
an
s
m
is
s
io
n
i
n
s
u
lato
r
s
ar
e
av
ailab
le
in
v
ar
io
u
s
t
y
p
es
an
d
s
h
ap
es,
w
h
ich
i
n
clu
d
e
i
n
d
iv
id
u
al
o
r
s
tr
in
g
s
o
f
d
is
k
s
,
lo
n
g
r
o
d
s
o
r
li
n
e
p
o
s
ts
.
T
h
er
e
ar
e
m
ai
n
l
y
th
r
ee
t
y
p
es
o
f
i
n
s
u
l
ato
r
s
u
s
ed
f
o
r
th
e
p
u
r
p
o
s
e
o
f
o
v
er
h
ea
d
in
s
u
lato
r
.
T
h
ey
ar
e
a.
P
i
n
I
n
s
u
la
to
r
,
b
.
Su
s
p
en
s
io
n
I
n
s
u
lato
r
an
d
c.
S
tr
ain
I
n
s
u
lato
r
.
T
h
er
e
ar
e
tw
o
m
o
r
e
t
y
p
es
o
f
elec
tr
ical
in
s
u
lato
r
s
w
h
ic
h
ar
e
av
ailab
le
m
ai
n
l
y
f
o
r
lo
w
v
o
lt
ag
e
ap
p
licatio
n
an
d
ar
e
ca
lled
Sta
y
I
n
s
u
la
to
r
an
d
Sh
ac
k
le
I
n
s
u
lato
r
.
T
h
e
in
s
u
la
to
r
s
ar
e
m
ad
e
o
f
g
lass
,
p
o
l
y
m
er
s
a
n
d
p
o
r
ce
lain
.
E
ac
h
m
o
d
el
is
m
ad
e
u
p
w
it
h
d
if
f
er
e
n
t
te
n
s
ile
s
tr
en
g
t
h
s
,
d
en
s
it
ies
a
n
d
d
if
f
er
e
n
t
le
v
els
o
f
p
er
f
o
r
m
an
ce
i
n
t
y
p
ical
wo
r
k
in
g
co
n
d
itio
n
s
.
C
er
a
m
ic
i
n
s
u
la
to
r
s
ar
e
g
en
er
all
y
u
s
ed
i
n
p
o
w
er
tr
an
s
m
is
s
i
o
n
an
d
d
is
tr
ib
u
t
io
n
li
n
es
f
o
r
a
lo
n
g
ti
m
e.
I
n
th
e
r
ec
en
t
ti
m
es,
p
o
l
y
m
er
ic
in
s
u
l
ato
r
s
h
a
v
e
b
ec
o
m
e
w
id
el
y
u
s
ef
u
l
d
u
e
t
h
eir
s
u
p
er
io
r
in
s
u
lat
io
n
p
er
f
o
r
m
a
n
ce
,
i
n
ter
m
s
o
f
co
n
ta
m
in
a
tio
n
e
n
d
u
r
a
n
ce
co
m
p
ar
ed
w
it
h
co
n
v
e
n
tio
n
al
ce
r
a
m
ic
i
n
s
u
lato
r
s
[
1
,
2
]
.
I
n
s
p
ec
tio
n
o
f
d
ef
ec
t
s
n
ee
d
s
to
b
e
d
o
n
e
f
o
r
a
w
id
e
v
ar
iet
y
o
f
ite
m
s
.
Gen
er
al
l
y
t
h
e
y
d
ep
en
d
o
n
i)
Size
o
f
th
e
ite
m
a
n
d
ii)
L
e
v
el
o
f
d
etails
r
eq
u
ir
ed
[
1
8
]
.
T
h
e
ty
p
es
o
f
ite
m
s
to
b
e
in
s
p
ec
ted
ar
e:
a.
L
ar
g
e
s
ca
le:
s
ag
g
i
n
g
s
p
a
n
s
,
b
r
o
k
en
o
r
s
lack
s
ta
y
w
ir
e
s
,
lean
i
n
g
p
o
les,
an
d
tr
ee
en
cr
o
ac
h
m
e
n
t;
b
.
Me
d
iu
m
s
ca
le:
eq
u
ip
m
e
n
t
f
i
x
ed
o
n
p
o
les,
air
b
r
ea
k
s
w
it
ch
es,
h
i
g
h
an
d
lo
w
v
o
lta
g
e
f
u
s
e
u
n
it
s
a
n
ti
-
cli
m
b
in
g
g
u
ar
d
s
,
an
d
s
a
f
et
y
n
o
tices
;
c.
S
m
al
l
s
ca
le:
c
h
ip
p
ed
o
r
b
r
o
k
en
in
s
u
lato
r
s
,
co
r
r
o
d
ed
jo
in
ts
o
n
co
n
d
u
cto
r
s
ca
u
s
i
n
g
d
is
co
l
o
r
atio
n
,
an
d
tr
ac
es
o
f
ar
cin
g
o
n
s
w
itc
h
es
o
r
f
u
s
e
g
ea
r
[
2
5
]
.
T
h
e
s
y
m
p
to
m
s
r
elate
d
to
ea
ch
p
r
o
b
le
m
o
n
o
v
er
h
ea
d
li
n
es
w
er
e
id
en
ti
fi
ed
a
n
d
q
u
a
n
ti
fi
ed
i
n
t
h
e
r
ep
o
r
t
o
f
E
P
R
I
[
3
2
]
.
Sin
ce
th
e
1
9
7
0
s
,
th
e
f
o
cu
s
o
n
cr
ac
k
i
n
g
o
f
i
n
s
u
lato
r
s
h
a
s
b
ee
n
o
b
s
er
v
ed
in
cr
ea
s
in
g
l
y
,
b
ec
au
s
e
s
a
f
et
y
i
s
s
u
es
r
elate
d
to
m
ec
h
a
n
ical
f
ac
to
r
s
h
av
e
d
ec
li
n
ed
[
4
]
.
L
o
o
m
s
h
a
s
b
r
ief
ed
[
5
2
]
th
at
th
e
d
am
a
g
e
o
f
p
in
an
d
ca
p
d
is
c
p
o
r
ce
lain
in
s
u
lato
r
s
is
m
ai
n
l
y
d
u
e
to
ce
m
e
n
t
g
r
o
w
th
,
c
y
clin
g
an
d
co
r
r
o
s
io
n
.
A
s
p
er
C
h
er
n
e
y
,
p
o
r
ce
lain
s
u
s
p
en
s
io
n
i
n
s
u
l
ato
r
f
ail
u
r
es
ar
e
d
u
e
a
v
o
l
u
m
e
ex
p
an
s
io
n
o
f
t
h
e
h
ar
d
en
ed
P
o
r
tlan
d
ce
m
e
n
t
g
r
o
u
t
in
t
h
e
p
in
h
o
le
o
f
t
h
e
in
s
u
la
to
r
w
h
ic
h
in
t
u
r
n
ca
u
s
e
s
r
ad
ial
cr
ac
k
s
[
5
4
]
in
th
e
p
o
r
ce
lain
s
u
s
p
en
s
io
n
i
n
s
u
lat
o
r
s
s
h
ell
s
.
W
h
e
n
t
h
er
e
is
s
u
f
f
icien
tl
y
lo
w
s
u
r
f
ac
e
r
e
s
is
ta
n
ce
,
th
er
e
w
ill
b
e
elo
n
g
atio
n
o
f
p
ar
tial
d
is
c
h
a
r
g
es
alo
n
g
t
h
e
i
n
s
u
lato
r
p
r
o
f
ile
w
h
ich
e
v
e
n
tu
a
ll
y
r
e
s
u
l
ts
in
f
la
s
h
o
v
er
o
f
in
s
u
lato
r
[
5
1
]
.
T
h
e
em
er
g
i
n
g
d
ef
ec
ts
f
o
r
m
ed
in
th
e
b
eg
i
n
n
i
n
g
s
ta
g
e
s
o
f
m
a
n
u
f
ac
t
u
r
in
g
p
r
o
ce
s
s
also
ar
e
th
e
ca
u
s
e
s
o
f
f
ail
u
r
e
o
f
i
n
s
u
lato
r
[
5
5
]
.
W
h
en
i
m
a
g
e
p
r
o
ce
s
s
in
g
t
ec
h
n
iq
u
es
ar
e
co
n
s
id
er
ed
f
o
r
au
to
m
at
ic
co
n
d
itio
n
m
o
n
ito
r
i
n
g
,
t
h
er
e
is
an
i
m
p
o
r
t
an
t
p
r
o
b
lem
o
f
p
r
ese
n
ce
o
f
b
ac
k
g
r
o
u
n
d
ele
m
e
n
t
s
o
n
th
e
i
m
a
g
es.
Us
u
all
y
f
o
r
th
e
i
m
a
g
es
tak
e
n
f
r
o
m
t
h
e
g
r
o
u
n
d
,
th
e
b
ac
k
g
r
o
u
n
d
is
q
u
ite
ea
s
y
to
r
e
m
o
v
e
s
k
y
w
it
h
clo
u
d
s
,
b
ec
au
s
e
o
f
th
e
h
ig
h
co
n
tr
ast.
St
ill,
t
h
er
e
m
a
y
b
e
a
s
it
u
atio
n
f
o
r
ae
r
ial
i
m
a
g
e
s
an
d
t
h
o
s
e
ta
k
e
n
f
r
o
m
t
h
e
g
r
o
u
n
d
,
w
h
er
e
t
h
e
b
ac
k
g
r
o
u
n
d
is
m
u
ch
m
o
r
e
c
o
m
p
le
x
li
k
e
p
r
ese
n
ce
o
f
tr
ee
s
,
b
u
ild
in
g
s
o
r
r
o
ad
s
.
B
o
th
ca
s
es
ar
e
tack
led
b
y
s
o
m
e
au
th
o
r
s
[
1
]
T
o
m
ee
t
t
h
e
g
r
o
w
i
n
g
d
e
m
a
n
d
f
o
r
r
eliab
le
en
er
g
y
s
u
p
p
l
y
,
ele
ctr
ic
u
tili
ties
ar
e
s
er
io
u
s
l
y
m
o
tiv
ated
to
d
r
o
p
m
ain
te
n
a
n
ce
co
s
ts
b
y
r
u
n
n
i
n
g
in
s
p
ec
tio
n
p
r
o
g
r
a
m
s
to
lo
ca
te,
f
o
llo
w
-
u
p
an
d
r
ep
air
an
y
s
u
b
s
ta
n
tial
f
ail
u
r
es
at
ea
r
lies
t
p
o
s
s
ib
le
s
t
ag
e
[
2
6
]
.
E
f
f
o
r
ts
o
n
r
esear
ch
f
o
cu
s
es
o
n
au
to
m
ati
n
g
t
h
e
p
o
w
er
li
n
e
in
s
p
ec
tio
n
p
r
o
ce
s
s
b
y
lo
o
k
i
n
g
f
o
r
ap
p
r
o
a
ch
es
t
h
at
f
u
l
f
il
l
th
e
v
ar
io
u
s
r
eq
u
ir
e
m
e
n
ts
o
f
th
e
i
n
s
p
ec
tio
n
[
2
8
]
.
T
h
e
w
e
llb
ein
g
an
al
y
s
is
o
f
p
r
o
v
id
in
g
a
m
ea
n
s
to
p
er
ce
iv
e
th
e
li
k
eli
h
o
o
d
o
f
th
e
s
y
s
te
m
’
s
ap
p
r
o
ac
h
in
g
h
az
a
r
d
o
u
s
s
tate
in
d
icate
s
w
h
e
th
er
t
h
e
in
s
u
la
to
r
d
is
tr
ib
u
tio
n
s
y
s
te
m
is
in
a
g
o
o
d
o
r
clo
s
e
to
a
r
is
k
y
s
tate
[
9
]
.
Var
io
u
s
m
eth
o
d
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
201
7
:
3
0
3
–
3
1
2
306
o
r
tech
n
iq
u
es
u
s
ed
f
o
r
m
o
n
ito
r
in
g
th
e
co
n
d
itio
n
a
n
d
clas
s
i
f
i
ca
tio
n
o
f
t
h
e
i
n
s
u
lato
r
s
a
n
d
t
h
e
an
al
y
s
i
s
d
o
n
e
as
p
er
th
e
liter
atu
r
e
ar
e
lis
ted
in
T
ab
le
1.
T
ab
le
1
.
L
is
t o
f
De
f
ec
t D
etec
ti
o
n
Me
th
o
d
s
M
e
t
h
o
d
R
e
f
e
r
e
n
c
e
A
n
a
l
y
si
s d
o
n
e
A
e
r
i
a
l
i
n
s
p
e
c
t
i
o
n
u
si
n
g
ma
n
n
e
d
h
e
l
i
c
o
p
t
e
r
[
1
5
,
1
8
]
P
r
o
d
u
c
e
d
a
t
h
e
o
r
e
t
i
c
a
l
a
n
d
e
x
p
e
r
i
me
n
t
a
l
b
a
si
s
f
o
r
sp
e
c
i
f
y
i
n
g
t
h
e
c
a
me
r
a
p
o
i
n
t
i
n
g
a
c
c
u
r
a
c
y
M
o
d
i
f
i
e
d
H
o
u
g
h
t
r
a
n
sf
o
r
m,
S
V
M
mo
d
e
l
C
o
mb
i
n
e
d
H
o
u
g
h
t
r
a
n
sf
o
r
m,
S
V
M
mo
d
e
l
D
O
S
T
-
A
N
F
I
S
C
o
mb
i
n
e
d
a
p
p
r
o
a
c
h
C
o
mb
i
n
e
d
W
a
v
e
l
e
t
,
H
M
M
W
a
v
e
l
e
t
-
S
V
M
C
o
mb
i
n
e
d
A
p
p
r
o
a
c
h
D
O
S
T
-
S
V
M
C
o
mb
i
n
e
d
a
p
p
r
o
a
c
h
W
i
e
n
e
r
f
i
l
t
e
r
i
n
g
,
F
C
M
,
C
o
n
n
e
c
t
e
d
c
o
mp
o
n
e
n
t
l
a
b
e
l
i
n
g
N
e
a
r
-
f
i
e
l
d
M
i
c
r
o
w
a
v
e
i
mag
i
n
g
,
A
N
N
[
3
6
]
[
3
7
]
[
1
]
[
3
4
]
M
o
d
i
f
i
e
d
H
o
u
g
h
t
r
a
n
sf
o
r
m
a
n
d
c
o
l
o
r
f
e
a
t
u
r
e
s a
r
e
e
x
t
r
a
c
t
e
d
a
n
d
S
V
M
i
s
u
se
d
f
o
r
c
l
a
ssi
f
i
c
a
t
i
o
n
W
a
v
e
l
e
t
t
r
a
n
sf
o
r
m i
s
u
se
d
f
o
r
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
a
n
d
S
V
M
i
s
u
se
d
f
o
r
c
l
a
ssi
f
i
c
a
t
i
o
n
D
i
scre
t
e
o
r
t
h
o
g
o
n
a
l
S
-
t
r
a
n
sf
o
r
m i
s
u
s
e
d
f
o
r
d
e
t
e
c
t
i
o
n
a
n
d
A
N
F
I
S
i
s u
se
d
f
o
r
c
l
a
ssi
f
i
c
a
t
i
o
n
T
e
mp
l
a
t
e
d
e
si
g
n
,
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
u
s
i
n
g
w
a
v
e
l
e
t
t
r
a
n
sf
o
r
m a
n
d
H
M
M
f
o
r
w
e
l
l
-
b
e
i
n
g
a
n
a
l
y
si
s o
f
i
n
su
l
a
t
o
r
s
[
9
]
W
a
v
e
l
e
t
f
e
a
t
u
r
e
s a
r
e
e
x
t
r
a
c
t
e
d
a
n
d
S
V
M
i
s
u
se
d
f
o
r
c
l
a
ssi
f
i
c
a
t
i
o
n
[
3
5
]
[
3
3
]
[
4
0
]
D
i
scre
t
e
o
r
t
h
o
g
o
n
a
l
S
-
t
r
a
n
sf
o
r
m i
s
u
s
e
d
f
o
r
d
e
t
e
c
t
i
o
n
a
n
d
S
V
M
i
s
u
se
d
f
o
r
c
l
a
ssi
f
i
c
a
t
i
o
n
I
n
su
l
a
t
o
r
i
mag
e
i
s
se
g
me
n
t
e
d
b
a
se
d
o
n
t
h
e
i
m
p
r
o
v
e
d
F
C
M
a
l
g
o
r
i
t
h
m
a
n
d
c
o
n
t
o
u
r
o
f
i
n
s
u
l
a
t
o
r
i
s l
a
b
e
l
e
d
b
y
u
s
i
n
g
c
o
n
n
e
c
t
e
d
c
o
m
p
o
n
e
n
t
l
a
b
e
l
i
n
g
a
l
g
o
r
i
t
h
m
A
n
o
v
e
l
a
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
(
A
N
N
)
-
b
a
se
d
n
e
a
r
-
f
i
e
l
d
mi
c
r
o
w
a
v
e
n
o
n
d
e
st
r
u
c
t
i
v
e
t
e
st
i
n
g
t
e
c
h
n
i
q
u
e
f
o
r
d
e
f
e
c
t
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
i
n
n
o
n
c
e
r
a
mi
c
i
n
s
u
l
a
t
o
r
s
T
h
e
u
lti
m
ate
o
b
j
ec
tiv
e
o
f
t
h
e
in
s
u
lato
r
d
etec
tio
n
s
y
s
te
m
is
to
ca
te
g
o
r
ize
d
ef
ec
ts
i
n
t
o
g
o
o
d
o
r
d
ef
ec
tiv
e
cla
s
s
es.
A
n
u
m
b
er
o
f
f
ea
t
u
r
es
h
a
v
e
to
b
e
ex
tr
ac
ted
f
r
o
m
r
eg
io
n
s
o
f
i
n
ter
est
a
n
d
id
ea
ll
y
,
d
if
f
er
e
n
t
co
m
b
i
n
atio
n
s
o
f
t
h
ese
f
ea
t
u
r
es
ar
e
r
eq
u
ir
ed
to
m
atc
h
u
n
iq
u
el
y
w
it
h
th
e
f
ea
t
u
r
es
o
f
d
if
f
er
en
t
t
y
p
es
o
f
d
ef
ec
ti
v
e
in
s
u
lato
r
s
.
T
h
e
v
ar
io
u
s
t
y
p
es
o
f
clas
s
i
f
icatio
n
alg
o
r
it
h
m
s
a
n
d
th
e
f
ea
tu
r
e
s
ex
tr
ac
ted
ar
e
s
h
o
w
n
i
n
t
h
e
T
ab
le
2
.
T
h
e
m
atc
h
i
n
g
is
g
e
n
er
all
y
d
o
n
e
u
s
i
n
g
ad
ap
ti
v
e
lear
n
i
n
g
m
et
h
o
d
s
s
u
p
er
v
is
ed
o
r
u
n
s
u
p
er
v
is
e
d
.
T
ab
le
2
.
L
is
t o
f
d
ef
ec
t c
la
s
s
i
f
i
ca
tio
n
m
et
h
o
d
s
P
a
p
e
r
M
e
t
h
o
d
o
f
C
l
a
ss
i
f
i
c
a
t
i
o
n
F
e
a
t
u
r
e
s e
x
t
r
a
c
t
e
d
[
3
6
]
[
9
]
S
V
M
S
V
M
R
G
B
c
o
l
o
r
f
e
a
t
u
r
e
s
S
t
a
t
i
st
i
c
a
l
f
e
a
t
u
r
e
s
[
3
4
]
H
M
M
S
t
a
t
i
st
i
c
a
l
f
e
a
t
u
r
e
s
[
1
]
A
N
F
I
S
D
O
S
T
f
e
a
t
u
r
e
s
[
3
5
]
A
N
F
I
S
&
S
V
M
M
e
a
n
,
V
a
r
i
a
n
c
e
5.
AUTOM
AT
I
C
I
N
SPEC
T
I
O
N
SYST
E
M
T
h
e
Dis
tr
ib
u
tio
n
Au
to
m
atio
n
S
y
s
te
m
(
D
AS)
w
as
d
ef
i
n
ed
b
y
I
n
s
tit
u
te
o
f
E
lectr
ical
an
d
E
lectr
o
n
ic
E
n
g
i
n
ee
r
s
(
I
E
E
E
)
as
a
s
y
s
te
m
w
h
ich
f
ac
ilit
ate
s
an
elec
tr
i
c
u
tili
t
y
to
co
o
r
d
in
ate,
m
o
n
it
o
r
an
d
o
p
er
ate
th
e
d
is
tr
ib
u
tio
n
co
m
p
o
n
e
n
ts
r
e
m
o
tel
y
,
f
r
o
m
r
e
m
o
te
lo
ca
tio
n
s
in
a
r
ea
l
t
i
m
e
m
o
d
e
[
4
8
]
.
T
h
e
DAS
i
n
v
o
lv
e
s
co
llectio
n
o
f
d
ata
an
d
a
n
al
y
z
e
th
e
in
f
o
r
m
at
io
n
to
m
a
k
e,
i
m
p
le
m
e
n
t
th
e
co
n
tr
o
l
d
ec
is
io
n
s
a
n
d
to
ac
h
ie
v
e
d
esire
d
r
esu
lt
[
4
9
]
.
T
h
e
Dis
tr
ib
u
tio
n
C
o
n
tr
o
l
C
en
tr
e
(
DC
C
)
is
th
e
lo
ca
tio
n
f
r
o
m
w
h
er
e
c
o
n
tr
o
l
d
ec
is
io
n
s
ar
e
i
n
itiated
.
I
n
d
a
y
to
d
a
y
o
p
e
r
atio
n
an
d
m
ai
n
ten
a
n
ce
o
f
p
o
w
er
d
is
tr
ib
u
tio
n
n
e
t
w
o
r
k
P
o
w
er
Di
s
tr
ib
u
t
io
n
Au
to
m
a
tio
n
is
a
d
ev
elo
p
in
g
f
i
eld
an
d
is
v
er
y
u
s
e
f
u
l
[
2
2
]
.
E
n
ac
t
m
e
n
t
o
f
elec
tr
icit
y
A
ct
2
0
0
3
o
f
I
n
d
ia
h
as
b
ee
n
p
r
ess
u
r
izi
n
g
th
e
p
o
w
er
d
is
tr
ib
u
tio
n
u
ti
lit
y
to
m
a
k
e
a
n
ice
p
lan
to
ac
h
iev
e
Dis
tr
ib
u
tio
n
A
u
to
m
at
io
n
w
i
th
o
u
t
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
R
ev
iew
o
f Ma
ch
in
e
V
is
io
n
b
a
s
ed
I
n
s
u
la
to
r
I
n
s
p
ec
tio
n
S
yste
ms fo
r
Ove
r
h
ea
d
.
..
(
P
.
S
u
r
ya
P
r
a
s
a
d
)
307
f
u
r
t
h
er
d
ela
y
[
4
7
]
.
T
h
e
au
to
m
atio
n
s
y
s
te
m
ca
n
b
e
d
esi
g
n
ed
u
s
in
g
t
h
e
av
ailab
le
r
eso
u
r
ce
s
li
k
e
co
m
p
u
ter
s
y
s
te
m
s
,
co
n
tr
o
l
s
y
s
te
m
s
a
n
d
m
eter
i
n
g
s
y
s
te
m
s
.
T
h
e
to
o
ls
s
u
c
h
as
co
m
p
u
ter
s
an
d
R
e
m
o
te
T
er
m
in
al
Un
i
ts
(
R
T
Us)
ar
e
u
s
ed
f
o
r
au
to
m
at
i
o
n
w
h
ic
h
u
s
e
s
u
ti
liti
e
s
to
co
n
tr
o
l
d
is
tr
ib
u
tio
n
s
y
s
te
m
s
a
n
d
en
h
a
n
ce
ef
f
icie
n
c
y
,
r
eliab
ilit
y
,
a
n
d
q
u
alit
y
o
f
elec
t
r
ic
s
er
v
ice
[
5
0
]
.
T
h
e
f
o
llo
w
in
g
ar
e
th
e
k
e
y
ele
m
en
t
s
o
f
th
e
au
to
m
atic
i
n
s
p
ec
tio
n
s
y
s
te
m
o
f
d
is
tr
ib
u
t
io
n
s
y
s
te
m
i
n
s
u
lato
r
s
.
5
.
1
.
I
m
a
g
e
Acquis
it
io
n
a
nd
P
ro
ce
s
s
ing
T
r
a
d
itio
n
all
y
t
h
e
cr
e
w
s
ar
e
s
e
n
t
o
u
t
b
y
t
h
e
u
t
ilit
ie
s
eit
h
er
o
n
f
o
o
t
o
r
in
v
e
h
icles
to
d
r
iv
e
th
e
li
n
es
,
w
it
h
f
r
eq
u
e
n
t
s
to
p
s
to
s
e
n
d
th
e
lin
e
m
e
n
cli
m
b
t
h
e
to
w
er
s
f
o
r
clo
s
er
in
s
p
ec
tio
n
[
4
1
]
.
A
n
al
ter
n
ati
v
e
ap
p
r
o
ac
h
is
to
tr
ai
n
i
n
s
p
ec
to
r
s
’
fly
ab
o
ar
d
th
e
h
e
lic
o
p
ter
s
in
i
n
s
p
ec
ti
n
g
th
e
li
n
es
w
ith
b
in
o
c
u
lar
s
a
n
d
ca
m
er
a
s
,
r
ec
o
r
d
th
e
d
ata
in
a
lo
g
b
o
o
k
as t
h
e
h
elico
p
ter
h
o
v
er
s
o
v
er
an
d
ar
o
u
n
d
p
o
w
er
lin
e
s
[
1
9
]
.
A
s
a
p
r
o
m
is
in
g
alter
n
a
tiv
e,
a
h
elico
p
ter
w
it
h
a
u
to
m
a
tic
v
i
d
eo
s
u
r
v
ei
llan
ce
s
y
s
te
m
w
as
s
u
g
g
ested
[
2
5
]
b
u
t
di
f
f
ic
u
lt
y
co
m
es
b
ec
a
u
s
e
o
f
r
o
tatio
n
o
f
t
h
e
ca
m
er
a
i
n
it
s
g
i
m
b
als
a
n
d
t
h
e
tr
an
s
latio
n
al
m
o
tio
n
o
f
th
e
h
elico
p
ter
[
2
0
]
.
A
n
i
n
n
o
v
ati
v
e
ap
p
r
o
ac
h
u
s
i
n
g
R
T
Us
a
s
w
ell
as
s
u
r
f
ac
e
v
eh
ic
u
lar
p
atr
o
lli
n
g
(
SVP
)
w
as
v
al
id
ated
b
y
d
i
f
f
er
en
t
r
esear
ch
er
s
.
W
ith
th
e
av
ailab
il
i
t
y
o
f
b
o
th
h
ig
h
r
eso
lu
t
io
n
v
id
eo
ca
m
er
a
s
an
d
co
m
m
u
n
icatio
n
n
e
t
w
o
r
k
s
f
o
r
m
a
s
s
u
s
e,
an
ar
r
an
g
e
m
en
t
o
f
i
n
telli
g
e
n
t
R
T
Us
f
itted
w
it
h
t
h
e
r
eq
u
ir
ed
eq
u
ip
m
e
n
t
ca
n
s
e
n
d
t
h
e
i
m
ag
e
s
o
f
th
e
in
s
u
la
to
r
s
to
th
e
co
n
tr
o
l
r
o
o
m
at
r
eg
u
lar
i
n
ter
v
al
f
o
r
f
u
r
th
er
a
n
al
y
s
i
s
[
1
,
3
5
]
.
Sin
ce
th
e
o
v
er
h
ea
d
d
i
s
tr
i
b
u
tio
n
n
et
w
o
r
k
s
i
n
I
n
d
ia
m
o
s
tl
y
r
u
n
p
ar
allel
to
th
e
r
o
ad
s
,
SVP
ap
p
r
o
ac
h
s
ee
m
s
to
b
e
a
v
iab
le
alter
n
ativ
e
[
9
,
3
4
,
3
6
,
3
7
]
to
th
e
ae
r
ial
in
s
p
ec
tio
n
m
et
h
o
d
u
s
i
n
g
h
el
ic
o
p
ter
s
,
w
h
ic
h
is
n
o
t o
n
l
y
co
s
tl
y
an
d
also
t
h
er
e
is
p
o
s
s
ib
ilit
y
o
f
d
an
g
er
.
5
.
2
.
E
x
t
ra
ct
io
n o
f
I
m
a
g
es P
er
t
a
i
nin
g
t
o
I
ns
ula
t
o
rs
T
r
a
d
itio
n
al
ap
p
r
o
ac
h
es
to
au
t
o
m
a
tic
p
o
w
er
lin
e
in
s
p
ec
tio
n
[
2
6
,
4
5
]
ar
e
b
ased
o
n
h
u
m
an
o
b
s
er
v
atio
n
an
d
ca
n
b
e
p
er
f
o
r
m
ed
s
i
m
p
l
y
f
r
o
m
th
e
g
r
o
u
n
d
,
w
i
th
cli
m
b
in
g
,
u
s
i
n
g
b
u
c
k
et
tr
u
c
k
m
e
th
o
d
s
o
r
an
air
b
o
r
n
e
p
latf
o
r
m
[
4
6
]
.
T
o
r
ed
u
ce
th
e
in
s
p
ec
tio
n
a
n
d
m
ai
n
te
n
an
ce
co
s
ts
,
n
e
w
ap
p
r
o
ac
h
es
b
ased
o
n
m
ac
h
i
n
e
v
i
s
io
n
tech
n
iq
u
es
w
it
h
an
al
y
s
is
o
f
v
i
d
eo
s
eq
u
en
ce
s
r
ec
o
r
d
ed
d
u
r
in
g
th
e
p
atr
o
l
h
av
e
b
ee
n
in
tr
o
d
u
ce
d
.
T
h
i
s
r
esu
lted
in
in
cr
ea
s
ed
r
o
b
u
s
t
n
es
s
o
f
t
h
e
p
o
w
er
s
y
s
te
m
a
n
d
also
i
s
h
el
p
f
u
l
in
t
h
e
a
u
to
m
ated
d
o
cu
m
en
tatio
n
.
Usa
g
e
o
f
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
e
s
is
ad
ap
tab
le
to
th
e
DSA
an
d
th
u
s
e
m
er
g
e
s
as
a
v
iab
le
o
p
tio
n
f
o
r
r
ea
l
-
ti
m
e
i
m
p
le
m
en
ta
tio
n
.
T
h
e
i
n
s
u
lato
r
m
o
n
ito
r
in
g
s
y
s
te
m
d
escr
ib
ed
is
d
ep
icted
in
t
h
e
b
lo
ck
d
ia
g
r
a
m
s
h
o
w
n
i
n
Fig
u
r
e
1
.
T
h
e
s
alien
t
f
ea
t
u
r
es
ar
e:
a.
i
m
a
g
e
ac
q
u
i
s
itio
n
co
n
ta
in
i
n
g
p
o
les
as
w
ell
a
s
in
s
u
la
to
r
s
,
b
.
ex
tr
ac
tio
n
o
f
i
m
a
g
e
pe
r
tain
in
g
to
in
s
u
la
to
r
s
an
d
c.
co
n
d
itio
n
m
o
n
ito
r
in
g
o
f
i
n
s
u
la
to
r
s
u
s
i
n
g
d
i
f
f
er
en
t
m
ac
h
in
e
le
ar
n
in
g
tec
h
n
iq
u
es.
Fig
u
r
e
1
.
I
n
s
u
la
to
r
Mo
n
ito
r
in
g
S
y
s
te
m
T
h
e
p
o
w
er
d
is
tr
ib
u
tio
n
s
y
s
te
m
cr
o
s
s
es
m
o
u
n
ta
in
s
a
n
d
f
o
r
ests
.
T
h
er
ef
o
r
e,
k
ee
p
in
g
i
n
v
ie
w
al
l
p
o
s
s
ib
le
s
itu
a
tio
n
s
,
p
lain
b
ac
k
g
r
o
u
n
d
a
n
d
co
m
p
lex
b
ac
k
g
r
o
u
n
d
i
m
a
g
es
h
av
e
b
ee
n
co
n
s
id
e
r
ed
as
test
ca
s
es
to
ex
tr
ac
t
in
s
u
lato
r
i
m
a
g
es
u
s
i
n
g
co
lo
r
f
ea
tu
r
es
a
n
d
w
a
v
ele
t
f
ea
tu
r
e
s
b
y
s
o
m
e
a
u
t
h
o
r
s
[
1
,
3
5
,
3
6
,
3
7
]
.
T
o
s
i
m
p
li
f
y
t
h
e
tas
k
o
f
ac
q
u
ir
i
n
g
th
e
i
m
a
g
es
co
n
tai
n
i
n
g
o
n
l
y
in
s
u
lato
r
s
,
i
m
ag
e
s
e
g
m
e
n
tatio
n
h
a
s
b
ee
n
s
u
cc
e
s
s
f
u
ll
y
d
o
n
e
u
s
i
n
g
p
r
o
v
en
tec
h
n
iq
u
es
s
u
c
h
a
s
C
a
n
n
y
ed
g
e
d
etec
tio
n
,
Ho
u
g
h
tr
a
n
s
f
o
r
m
in
co
n
j
u
n
ctio
n
w
it
h
SVM
[
3
7
]
.
I
n
ea
c
h
b
o
u
n
d
in
g
b
o
x
,
th
e
i
n
s
u
lato
r
’
s
p
r
ese
n
ce
w
a
s
d
etec
ted
b
y
ex
tr
ac
ti
n
g
s
o
m
e
f
ea
t
u
r
es
l
ik
e
co
lo
r
f
ea
tu
r
es
[
3
3
]
an
d
e
x
tr
ac
ted
f
ea
t
u
r
es
f
r
o
m
t
h
e
b
o
u
n
d
i
n
g
b
o
x
es
w
er
e
s
u
p
p
lied
to
an
SVM
clas
s
i
f
ier
a
n
d
p
r
esen
ce
o
f
in
s
u
lato
r
w
a
s
d
etec
ted
.
1
1
k
V
Su
b
s
t
a
t
i
o
n
2
2
0
k
V
S
u
b
s
t
a
t
i
o
n
RT
U
R
T
U
U
t
i
l
i
z
a
t
i
on
M
O
D
EM
3
3
k
V
S
u
b
s
t
a
t
i
o
n
P
r
o
t
e
c
t
i
o
n
M
a
i
n
t
e
n
a
n
c
e
C
e
n
t
e
r
S
u
b
st
a
t
i
o
n
C
o
n
t
r
o
l
S
y
st
e
m
S
M
S
w
i
t
h
A
r
c
h
i
v
i
n
g
Ev
a
l
u
a
t
i
o
n
M
O
D
EM
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
201
7
:
3
0
3
–
3
1
2
308
6.
DIS
CU
SS
I
O
N
O
N
D
E
F
E
CT
DE
T
E
C
T
I
O
N
M
E
T
H
O
DS
T
h
e
in
s
u
la
to
r
s
o
f
a
p
o
w
er
d
is
tr
ib
u
tio
n
s
y
s
te
m
ar
e
m
o
n
ito
r
ed
u
s
in
g
a
d
ef
ec
t
d
et
ec
tio
n
an
d
class
i
f
icatio
n
in
t
h
r
ee
s
tep
s
p
r
o
ce
d
u
r
e:
Seg
m
e
n
tat
io
n
lo
ca
liz
e
ca
n
d
id
ate
d
ef
ec
ts
o
r
r
eg
io
n
s
o
f
I
n
ter
est
(
R
o
I
)
,
f
ea
t
u
r
e
ex
tr
ac
tio
n
f
r
o
m
R
o
I
s
an
d
last
l
y
,
g
o
o
d
o
r
d
ef
ec
tiv
e
class
i
f
icatio
n
.
T
h
e
s
e
g
m
e
n
ta
tio
n
h
a
s
b
ee
n
v
er
y
s
u
cc
e
s
s
f
u
l
in
[
1
,
9
,
3
3
,
3
4
,
3
5
,
3
6
,
3
7
]
b
y
th
e
u
s
e
o
f
v
ar
io
u
s
t
y
p
es
o
f
tech
n
iq
u
es.
I
n
t
h
e
ca
s
e
o
f
m
i
n
o
r
cr
ac
k
s
o
n
th
e
in
s
u
la
to
r
s
,
th
e
d
etec
ti
o
n
alo
n
e
i
s
a
v
er
y
u
s
e
f
u
l
t
o
o
l
to
id
en
tify
t
h
e
d
ef
ec
t
a
n
d
p
er
f
ec
t
le
v
el
o
f
class
i
f
icatio
n
is
ac
h
iev
ed
w
it
h
ea
s
e.
6
.
1
.
Sp
a
t
ia
l
Do
m
a
in
-
ba
s
ed
M
et
ho
ds
T
h
er
e
is
p
len
t
y
o
f
liter
atu
r
e
a
v
ailab
le
f
o
r
au
to
m
atic
d
etec
ti
o
n
o
f
elec
tr
ic
to
w
er
s
a
n
d
f
e
w
r
ef
er
en
ce
s
f
o
r
ex
tr
ac
tio
n
a
n
d
clas
s
i
f
icati
o
n
o
f
i
n
s
u
lato
r
s
.
D
u
tta,
T
an
im
a,
et
al
[
4
3
]
h
av
e
ta
k
e
n
ae
r
ial
i
m
a
g
er
y
u
s
i
n
g
Un
m
an
n
ed
A
er
ial
Veh
icle
(
U
A
V)
w
it
h
v
ar
y
in
g
n
a
tu
r
al
a
n
d
co
m
p
le
x
s
u
r
r
o
u
n
d
i
n
g
s
to
d
etec
t
lin
e
f
a
u
lt
s
u
s
i
n
g
a
n
o
v
el
m
o
r
p
h
o
lo
g
ical
o
p
er
ato
r
,
an
d
a
r
o
b
u
s
t
i
m
ag
e
s
p
ac
e
h
eu
r
i
s
tics
to
lo
ca
te
an
d
e
x
tr
ac
t
p
o
w
er
li
n
es
co
m
p
lete
l
y
.
Ma
r
ti
n
ez
,
et
al.
p
r
esen
t
s
a
n
ap
p
r
o
ac
h
[
4
4
]
f
o
cu
s
i
n
g
o
n
a
u
to
n
o
m
o
u
s
d
etec
tio
n
i
n
r
ea
l
-
ti
m
e,
elec
tr
ic
to
w
er
s
lo
ca
lizatio
n
an
d
tr
ac
k
i
n
g
u
s
i
n
g
a
s
tr
ateg
y
to
tr
ain
a
t
w
o
-
clas
s
m
u
ltil
a
y
er
p
er
ce
p
tr
o
n
(
ML
P
)
n
eu
r
al
n
et
w
o
r
k
a
n
d
ap
p
lied
o
v
er
s
l
id
in
g
w
i
n
d
o
w
s
f
o
r
ea
ch
ca
m
er
a
f
r
a
m
e
u
n
ti
l
a
to
w
er
i
s
d
etec
ted
.
E
lectr
ic
to
w
er
s
a
s
w
ell
as
t
h
e
in
s
u
la
to
r
s
ar
e
ex
tr
ac
ted
b
y
V.
S.M
u
r
t
h
y
et
al.
[
3
6
,
3
7
]
b
y
co
n
v
er
tin
g
th
e
co
lo
u
r
i
m
a
g
es
f
ir
s
t
in
to
g
r
e
y
s
ca
led
i
m
a
g
es
a
n
d
th
e
n
C
an
n
y
ed
g
e
d
etec
tio
n
w
h
ic
h
is
f
o
llo
w
ed
b
y
th
e
m
o
d
if
ied
H
o
u
g
h
tr
an
s
f
o
r
m
.
I
t
w
a
s
u
s
ed
to
i
s
o
late
f
ea
tu
r
e
s
o
f
a
p
ar
tic
u
lar
s
h
ap
e
w
it
h
i
n
a
n
i
m
a
g
e
a
n
d
s
e
g
m
en
tatio
n
is
r
ea
lized
b
y
ap
p
l
y
i
n
g
ed
g
e
d
etec
tio
n
an
d
th
r
e
s
h
o
ld
i
n
g
m
et
h
o
d
s
.
I
t
is
also
u
s
ed
as
a
to
o
l
f
o
r
ed
g
e
lin
k
i
n
g
to
o
.
I
n
an
o
th
er
tech
n
iq
u
e,
K
-
m
ea
n
s
cl
u
s
ter
i
n
g
h
as
b
ee
n
u
s
ed
to
lo
ca
te
t
h
e
p
r
o
p
er
b
o
u
n
d
in
g
b
o
x
es
co
n
tai
n
i
n
g
th
e
in
s
u
lato
r
s
.
Fro
m
t
h
e
b
o
u
n
d
in
g
b
o
x
es,
f
ea
t
u
r
es
li
k
e
m
ea
n
,
s
tan
d
ar
d
d
ev
iatio
n
w
er
e
ex
tr
ac
ted
an
d
s
u
cc
es
s
f
u
ll
y
t
ested
f
o
r
au
to
m
atic
in
s
u
lato
r
ex
tr
ac
tio
n
f
r
o
m
p
la
in
b
ac
k
g
r
o
u
n
d
[
3
5
]
an
d
co
m
p
lex
b
ac
k
g
r
o
u
n
d
[
1
]
as
well.
Fu
zz
y
C
-
Me
an
s
alg
o
r
ith
m
(
F
C
M)
as
p
r
o
p
o
s
ed
b
y
B
o
W
en
W
an
g
,
Q
u
an
Gu
[
3
3
]
is
u
s
ed
to
r
ec
o
g
n
iz
e
tr
an
s
m
is
s
io
n
lin
e
in
s
u
lato
r
s
.
T
o
f
iltra
te
an
d
r
ec
o
v
er
i
m
a
g
e
in
p
r
e
-
p
r
o
ce
s
s
i
n
g
,
th
e
i
m
p
r
o
v
ed
W
ien
er
f
ilter
alg
o
r
ith
m
w
as
u
s
e
d
an
d
th
e
n
,
i
m
p
r
o
v
ed
FC
M
w
a
s
u
s
ed
to
s
eg
m
e
n
t
t
h
e
in
s
u
la
t
o
r
.
Fin
all
y
,
t
h
e
co
n
to
u
r
o
f
i
n
s
u
lato
r
is
lab
eled
b
y
u
s
i
n
g
co
n
n
ec
ted
co
m
p
o
n
e
n
t
la
b
elin
g
al
g
o
r
ith
m
.
So
m
e
e
f
f
ici
en
t
alg
o
r
it
h
m
s
s
u
c
h
as
te
m
p
la
te
d
esig
n
a
n
d
m
ea
n
s
h
i
f
t
tr
ac
k
in
g
h
av
e
s
u
cc
e
s
s
f
u
ll
y
tr
ac
k
ed
t
h
e
p
o
les
in
th
e
s
tr
ea
m
i
n
g
v
id
eo
f
o
r
t
h
e
s
u
b
s
eq
u
en
t
p
r
o
ce
s
s
o
f
id
en
ti
f
y
i
n
g
in
s
u
lato
r
s
[
3
4
]
.
A
m
o
n
g
t
h
e
clu
s
ter
in
g
m
et
h
o
d
s
u
s
ed
to
s
eg
m
e
n
t
t
h
e
o
b
j
ec
t
o
f
in
ter
es
t,
th
e
K
-
Me
an
s
al
g
o
r
ith
m
i
n
v
o
lv
e
s
m
o
r
e
er
r
o
r
clu
s
ter
p
ix
el
s
,
w
h
er
ea
s
FC
M
a
lg
o
r
it
h
m
in
v
o
l
v
es
less
.
T
h
e
alg
o
r
it
h
m
u
s
i
n
g
FC
M
[
3
3
]
h
av
e
a
g
o
o
d
s
eg
m
e
n
tat
io
n
e
f
f
ec
t,
e
f
f
ec
ti
v
el
y
r
ed
u
cin
g
t
h
e
n
u
m
b
er
o
f
er
r
o
r
clu
s
ter
p
ix
e
ls
.
6
.
2
.
F
re
qu
ency
Do
m
a
in
-
ba
s
ed
M
et
ho
ds
I
n
th
e
r
ep
o
r
ted
r
esear
ch
o
n
co
n
d
itio
n
m
o
n
ito
r
in
g
o
f
th
e
in
s
u
lato
r
s
,
th
e
d
eg
r
ee
o
f
d
a
m
ag
e
o
f
an
in
s
u
lato
r
af
f
ec
ts
t
h
e
d
is
tr
ib
u
tio
n
s
y
s
te
m
in
d
if
f
er
e
n
t
w
a
y
s
an
d
s
o
th
e
d
ef
ec
t
o
f
in
s
u
lato
r
s
h
a
v
e
b
ee
n
ca
teg
o
r
ized
in
to
eit
h
er
t
h
r
ee
s
tates
n
a
m
e
l
y
g
o
o
d
,
m
ar
g
in
al
an
d
r
is
k
s
ta
tes
o
r
t
w
o
s
tates,
g
o
o
d
an
d
r
is
k
y
.
T
o
u
n
d
er
s
ta
n
d
th
e
co
n
d
itio
n
o
f
in
s
u
lato
r
s
f
r
o
m
t
h
e
ac
q
u
ir
ed
im
ag
e
s
,
th
e
ad
o
p
ted
f
ea
tu
r
es
e
x
tr
ac
ted
b
y
a
u
t
h
o
r
s
u
s
e
s
f
a
m
ilie
s
o
f
w
a
v
elet
tr
an
s
f
o
r
m
[
3
4
,
3
6
,
3
7
]
as w
e
ll a
s
d
is
c
r
ete
o
r
th
o
g
o
n
al
S
-
tr
an
s
f
o
r
m
(
DOST
)
[
1
,
3
5
]
.
I
f
th
e
o
b
j
ec
t’
s
s
ize
is
s
m
all
o
r
th
er
e
is
lo
w
co
n
tr
ast,
g
e
n
er
all
y
th
e
y
ar
e
to
b
e
ex
am
in
ed
at
h
ig
h
re
s
o
lu
tio
n
s
.
I
f
th
e
ir
s
ize
is
lar
g
e
o
r
co
n
tr
ast
is
h
i
g
h
,
a
co
ar
s
e
v
ie
w
is
n
ee
d
ed
.
I
f
b
o
th
s
m
a
ll
an
d
lar
g
e
o
b
j
ec
ts
,
as
w
el
l
as
lo
w
a
n
d
h
ig
h
co
n
tr
ast
o
b
j
ec
ts
ar
e
s
i
m
u
ltan
eo
u
s
l
y
p
r
esen
t,
it
is
b
en
ef
icia
l
to
s
tu
d
y
t
h
e
m
at
s
ev
er
al
r
eso
lu
tio
n
s
.
T
h
is
co
n
ce
p
t
is
th
e
f
u
n
d
a
m
en
tal
m
o
tiv
a
tio
n
f
o
r
m
u
lti
-
r
eso
l
u
tio
n
an
a
l
y
s
is
(
MR
A
)
[
3
0
]
.
T
h
e
w
a
v
elet
tr
an
s
f
o
r
m
[
6
9
]
h
as
t
h
e
n
at
u
r
al
ab
ilit
y
i
n
ca
p
tu
r
in
g
ev
e
n
s
m
a
ll
cr
ac
k
s
o
n
th
e
i
n
s
u
lato
r
u
s
i
n
g
MR
A
alo
n
g
w
it
h
a
class
i
f
y
in
g
tech
n
iq
u
e
li
k
e
SVM
to
d
is
ce
r
n
th
e
h
ea
lt
h
o
f
th
e
i
n
s
u
lato
r
.
T
h
e
DOST
is
u
s
ed
in
f
i
n
d
in
g
th
e
co
n
d
itio
n
[
5
6
,
5
7
]
o
f
in
s
u
lato
r
f
r
o
m
t
h
e
ex
tr
ac
ted
i
m
a
g
es.
I
t
g
i
v
es
a
s
p
atial
f
r
eq
u
en
c
y
r
ep
r
esen
tatio
n
li
k
e
DW
T
.
I
t
also
h
as
a
n
ad
d
itio
n
al
b
en
e
f
it
th
at
p
h
ase
p
r
o
p
er
ties
o
f
t
h
e
ST
an
d
FT
ar
e
m
ai
n
tai
n
ed
an
d
r
etain
s
t
h
e
ab
il
it
y
to
co
llap
s
e
ex
ac
tl
y
b
ac
k
to
th
e
Fo
u
r
ier
d
o
m
ai
n
[
5
8
]
.
7.
DE
F
E
C
T
C
L
AS
SI
F
I
C
AT
I
O
N
SYST
E
M
S
7
.
1
.
Dis
cu
s
s
io
n o
n
Def
ec
t
Cla
s
s
if
i
ca
t
io
n
M
et
ho
ds
On
ce
t
h
e
d
etec
tio
n
o
f
i
n
s
u
la
to
r
is
d
o
n
e,
clas
s
if
icatio
n
o
f
d
ef
ec
ts
i
n
elec
tr
ical
p
o
w
er
s
y
s
te
m
i
n
s
u
lato
r
s
to
id
en
tify
an
d
s
u
b
s
eq
u
e
n
tl
y
r
ep
lacin
g
t
h
e
f
a
u
lt
y
i
n
s
u
lato
r
s
h
en
ce
p
r
ev
e
n
ti
n
g
h
ea
v
y
d
a
m
a
g
es.
T
h
e
in
tell
ig
e
n
t
tech
n
iq
u
es
s
u
ch
as
S
VM
,
H
MM
(
Hid
d
en
Ma
r
k
o
v
Mo
d
el)
an
d
A
NFI
S
(
A
d
ap
ti
v
e
Ne
u
r
o
Fu
zz
y
I
n
f
er
en
ce
S
y
s
te
m
)
h
a
v
e
u
s
ed
s
u
c
h
f
ea
t
u
r
es
to
p
er
f
o
r
m
t
h
e
tas
k
o
f
au
to
m
ated
co
n
d
itio
n
m
o
n
ito
r
in
g
o
f
t
h
e
i
n
s
u
lato
r
ef
f
icien
tl
y
.
A
s
SVM
o
p
er
ates
o
n
s
tr
u
ctu
r
al
r
is
k
m
i
n
i
m
izatio
n
p
r
in
cip
le
in
m
in
i
m
iz
in
g
a
n
u
p
p
er
b
o
u
n
d
b
ased
o
n
th
e
g
en
er
aliza
tio
n
er
r
o
r
,
an
d
d
ea
ls
w
it
h
o
n
l
y
t
w
o
p
ar
a
m
e
ter
s
,
f
o
r
class
i
f
icatio
n
an
d
it
h
as
b
ee
n
ch
o
s
e
n
f
o
r
ef
f
ec
tiv
e,
f
a
s
ter
i
n
s
u
lato
r
co
n
d
itio
n
m
o
n
i
to
r
in
g
f
o
r
au
to
m
ati
o
n
p
u
r
p
o
s
es
b
y
t
h
e
r
esear
c
h
er
s
[
1
]
,
[
9
]
.
A
NFI
S
is
an
ad
ap
tiv
e
n
et
w
o
r
k
w
h
ic
h
is
s
i
m
ilar
to
ad
ap
tiv
e
n
et
w
o
r
k
s
i
m
u
lato
r
o
f
f
u
zz
y
co
n
tr
o
ller
s
a
n
d
it is
eq
u
iv
a
len
t
to
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
R
ev
iew
o
f Ma
ch
in
e
V
is
io
n
b
a
s
ed
I
n
s
u
la
to
r
I
n
s
p
ec
tio
n
S
yste
ms fo
r
Ove
r
h
ea
d
.
..
(
P
.
S
u
r
ya
P
r
a
s
a
d
)
309
a
FIS.
B
ac
k
p
r
o
p
ag
atio
n
g
r
a
d
ien
t
d
escen
t
an
d
leas
t
s
q
u
ar
e
m
et
h
o
d
f
o
r
n
o
n
-
li
n
ea
r
an
d
lin
ea
r
p
ar
a
m
eter
s
i
s
u
s
ed
to
ad
j
u
s
t
th
e
p
ar
a
m
e
ter
s
f
o
r
a
g
iv
e
n
i
n
p
u
t
o
r
o
u
tp
u
t
d
ata
s
et
[
8
]
.
A
co
m
p
r
eh
e
n
s
i
v
e
r
ev
ie
w
o
f
co
m
m
o
n
m
ac
h
in
e
lear
n
i
n
g
tech
n
iq
u
e
s
l
ik
e
A
NN,
SVM,
an
d
GM
M
al
o
n
g
w
i
th
H
MM
th
a
t
a
r
e
u
s
ed
i
n
au
to
m
atic
s
p
ee
ch
r
ec
o
g
n
itio
n
is
g
iv
e
n
i
n
[
6
1
]
.
7
.
2
.
Su
pp
o
rt
Vec
t
o
r
M
a
chines
(
S
VM
)
SVM
is
a
s
et
o
f
s
u
p
er
v
i
s
ed
le
ar
n
in
g
al
g
o
r
ith
m
w
h
ic
h
ca
n
b
e
u
s
ed
f
o
r
class
if
icatio
n
a
n
d
r
eg
r
ess
io
n
[
6
4
]
.
I
t
is
f
o
u
n
d
as
a
v
er
y
ef
f
e
ctiv
e
tec
h
n
iq
u
e
f
o
r
g
en
er
al
p
u
r
p
o
s
e
s
u
p
er
v
is
ed
p
atter
n
r
ec
o
g
n
i
tio
n
as
p
r
o
p
o
s
ed
b
y
Vap
n
ik
et
al.
[
6
7
]
.
I
t
is
a
b
in
ar
y
n
o
n
-
li
n
ea
r
class
i
f
ier
e
m
p
l
o
y
ed
to
p
r
ed
ict
w
h
et
h
er
an
i
n
p
u
t
v
al
u
e
b
elo
n
g
s
a
class
1
o
r
a
class
2
an
d
is
g
en
er
all
y
u
s
ed
to
class
i
f
y
a
n
o
b
j
ec
t
in
to
a
d
ef
ec
tiv
e
o
r
g
o
o
d
o
n
e.
T
o
class
if
y
th
e
in
s
u
lato
r
s
,
SVM
s
ca
n
b
e
u
s
ed
to
s
ep
ar
ate
th
e
g
iv
en
s
e
t
o
f
lab
eled
d
ata
w
it
h
a
h
y
p
er
p
lan
e
w
it
h
t
h
e
m
a
x
i
m
u
m
m
ar
g
i
n
.
As
m
o
s
t
o
f
t
h
e
p
r
ac
ti
ca
l
class
i
f
ica
tio
n
p
r
o
b
le
m
s
ar
e
n
o
n
-
li
n
ea
r
,
t
h
e
SVM
s
u
s
e
k
er
n
el
f
u
n
ctio
n
s
t
h
at
au
to
m
at
icall
y
r
ea
l
ize
a
n
o
n
-
li
n
ea
r
m
ap
p
in
g
to
a
f
ea
t
u
r
e
s
p
a
ce
.
S
h
ap
e
r
ec
o
g
n
it
io
n
o
f
t
y
r
e
m
ar
k
i
n
g
p
o
in
ts
[
6
5
]
is
a
r
ec
en
t a
p
p
licatio
n
f
o
r
SV
M.
SVM
is
a
ls
o
ex
ten
d
ed
to
s
o
l
v
e
m
u
l
ticlas
s
s
ep
ar
atio
n
p
r
o
b
le
m
w
h
ic
h
u
s
es
o
n
e
-
v
er
s
u
s
-
al
l
an
d
o
n
e
-
v
er
s
u
s
-
o
n
e
tech
n
iq
u
es.
Mu
lti
class
p
r
o
b
lem
s
o
l
u
tio
n
u
s
in
g
SVM
h
as
b
ee
n
r
ep
o
r
ted
in
[
3
5
,
3
6
,
3
7
]
.
A
f
e
w
b
in
ar
y
clas
s
i
f
ier
s
ar
e
n
ee
d
ed
to
b
e
tr
ain
ed
to
f
o
r
a
m
u
lticl
ass
clas
s
i
f
icatio
n
p
r
o
b
le
m
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
SVM
clas
s
i
f
icatio
n
is
s
tr
o
n
g
l
y
r
el
ated
to
th
e
c
h
o
ice
o
f
t
h
e
k
er
n
el
f
u
n
ctio
n
.
T
h
er
e
ar
e
m
an
y
n
u
m
b
er
o
f
k
er
n
el
f
u
n
ctio
n
s
a
v
ailab
le
as
f
o
llo
w
s
:
lin
ea
r
k
er
n
el
f
u
n
ct
io
n
,
r
ad
ial
b
asis
f
u
n
ctio
n
,
p
o
l
y
n
o
m
ial
k
er
n
el
f
u
n
ctio
n
et
a
l.
Am
o
n
g
t
h
e
m
,
R
ad
ial
B
asis
F
u
n
ct
io
n
(
R
B
F)
is
t
h
e
m
o
s
t
p
o
p
u
lar
o
n
e.
T
h
e
s
am
p
le
s
ar
e
n
o
n
-
li
n
ea
r
l
y
m
ap
p
ed
in
to
a
h
i
g
h
er
d
i
m
e
n
s
io
n
al
s
p
a
ce
.
T
h
e
lin
ea
r
k
er
n
el
i
s
a
s
p
ec
ial
ca
s
e
o
f
R
B
F
[
6
8
]
.
T
h
e
f
ea
tu
r
es
ex
tr
ac
ted
f
r
o
m
DOST
,
alo
n
g
w
it
h
SVM
a
n
d
A
N
FIS
[
1
]
w
er
e
u
s
ed
to
es
ti
m
ate
th
e
co
n
d
itio
n
o
f
t
h
e
i
n
s
u
lato
r
.
I
n
t
h
e
r
e
f
er
r
ed
p
ap
er
s
[
1
,
3
5
]
SVM
is
u
s
ed
f
o
r
t
w
o
p
u
r
p
o
s
es,
i.e
.
f
o
r
lo
c
atin
g
t
h
e
p
r
o
p
er
b
o
u
n
d
in
g
b
o
x
es
co
n
tain
in
g
t
h
e
in
s
u
lato
r
s
a
m
o
n
g
s
t
th
e
b
o
u
n
d
in
g
b
o
x
es
u
s
i
n
g
f
ea
tu
r
e
s
(
lik
e
m
ea
n
a
n
d
s
tan
d
ar
d
d
ev
iatio
n
)
ex
tr
ac
ted
f
r
o
m
th
e
cr
o
p
p
e
d
im
a
g
es a
f
ter
ap
p
l
y
i
n
g
s
o
m
e
s
e
g
m
en
ta
tio
n
tec
h
n
i
q
u
e
an
d
also
f
o
r
class
i
f
y
i
n
g
i
n
s
u
la
to
r
ac
co
r
d
in
g
to
its
co
n
d
itio
n
.
So
,
th
e
o
u
tp
u
t
o
f
f
ir
s
t
SVM
g
i
v
es
b
o
u
n
d
i
n
g
b
o
x
es
h
a
v
in
g
in
s
u
lato
r
s
an
d
s
ec
o
n
d
SVM
d
is
cr
i
m
i
n
ates,
w
h
et
h
er
th
e
i
n
s
u
lato
r
is
h
ea
l
th
y
o
r
b
r
o
k
en
.
7
.
3
.
H
idd
en
M
a
rk
o
v
M
o
del (
H
M
M
)
HM
M
is
a
s
tat
is
tical
Ma
r
k
o
v
m
o
d
el
co
n
s
is
tin
g
o
f
a
Ma
r
k
o
v
ch
ai
n
w
it
h
f
in
i
te
n
u
m
b
er
o
f
s
tates,
a
s
tate
tr
an
s
itio
n
p
r
o
b
ab
ilit
y
m
atr
i
x
,
an
d
an
i
n
itial
s
tate
p
r
o
b
a
b
ilit
y
d
is
tr
ib
u
t
io
n
.
A
lt
h
o
u
g
h
th
e
s
tates
ar
e
u
n
o
b
s
er
v
ab
le,
b
u
t
th
e
o
u
tp
u
t
w
h
ic
h
i
s
d
ep
en
d
en
t
o
n
th
e
s
t
ate,
is
v
i
s
ib
le
a
n
d
ar
e
d
r
a
w
n
as
p
er
a
p
r
o
b
ab
ilit
y
d
is
tr
ib
u
tio
n
[
5
9
]
.
HM
Ms
w
er
e
u
s
ed
f
o
r
f
ac
e
d
etec
tio
n
an
d
r
ec
o
g
n
itio
n
an
d
th
e
y
w
er
e
m
o
ti
v
ated
b
ec
au
s
e
th
e
y
ar
e
p
ar
tially
i
n
v
ar
ian
t
to
v
ar
ia
tio
n
s
i
n
s
ca
li
n
g
an
d
also
th
e
s
tr
u
ctu
r
e
o
f
i
m
a
g
es
[
6
0
-
6
2
]
,
[
7
2
]
.
A
p
p
licatio
n
o
f
HM
Ms i
s
d
o
n
e
i
n
m
a
n
y
f
ie
ld
s
w
h
er
e
t
h
e
g
o
al
is
to
r
ec
o
v
er
a
d
ata
s
eq
u
en
ce
t
h
at
is
n
o
t o
b
s
er
v
ab
le
i
m
m
ed
iatel
y
w
h
er
ea
s
s
o
m
e
o
th
er
d
ata
ar
e
a
v
ailab
le
t
h
at
d
ep
en
d
o
n
t
h
e
s
e
q
u
en
ce
.
I
t is b
ei
n
g
u
s
ed
in
s
ev
er
al
f
ield
s
i
n
cl
u
d
in
g
f
ac
ial
ex
p
r
es
s
io
n
r
ec
o
g
n
itio
n
(
FER)
[
6
3
]
w
it
h
i
m
p
r
o
v
ed
ac
cu
r
ac
y
t
h
an
e
x
is
t
in
g
m
e
th
o
d
s
.
An
alg
o
r
it
h
m
g
i
v
e
n
in
[
3
4
]
u
tili
ze
s
h
id
d
en
Ma
r
k
o
v
m
o
d
els
to
d
eter
m
i
n
e
th
e
h
e
alth
co
n
d
itio
n
o
f
in
s
u
la
to
r
s
.
Sin
ce
in
s
u
lato
r
s
o
n
p
o
les
p
o
s
s
ess
a
lo
t
o
f
v
ar
iat
io
n
s
in
ter
m
s
o
f
f
ea
t
u
r
es,
a
n
d
al
s
o
r
eq
u
ir
es
s
ca
li
n
g
,
HM
M
h
a
s
b
ee
n
e
m
p
lo
y
ed
b
y
th
o
s
e
au
t
h
o
r
s
.
HM
M
h
as
b
ee
n
u
s
ed
f
o
r
w
ell
-
b
ein
g
an
al
y
s
i
s
to
s
eg
r
eg
ate
a
g
o
o
d
in
s
u
lat
o
r
f
r
o
m
a
b
ad
o
n
e
u
s
i
n
g
B
au
m
-
W
elc
h
alg
o
r
it
h
m
.
7
.
4
.
Ada
ptiv
e
Neuro
-
f
uzzy
I
nfe
re
nce
Sy
s
t
e
m
(
ANF
I
S)
Fo
r
a
g
iv
e
n
i
n
p
u
t
o
r
o
u
tp
u
t
d
a
ta
s
et,
th
e
ANFI
S
ad
j
u
s
t
s
all
t
h
e
r
eq
u
ir
ed
p
ar
am
e
ter
s
.
I
t
u
s
e
s
a
m
et
h
o
d
o
f
b
ac
k
p
r
o
p
ag
atio
n
g
r
ad
ien
t
d
escen
t
f
o
r
n
o
n
-
l
in
ea
r
p
ar
a
m
eter
s
an
d
lea
s
t
s
q
u
ar
e
t
y
p
e
o
f
m
et
h
o
d
f
o
r
li
n
ea
r
p
ar
am
eter
s
[
7
0
]
.
T
h
e
n
eu
r
o
-
a
d
ap
tiv
e
lear
n
i
n
g
tec
h
n
iq
u
es
p
r
o
v
id
e
a
m
et
h
o
d
o
lo
g
y
f
o
r
t
h
e
f
u
zz
y
m
o
d
elin
g
p
r
o
ce
d
u
r
e
in
ex
tr
ac
ti
n
g
in
f
o
r
m
atio
n
ab
o
u
t
a
d
ata
s
e
t.
T
h
is
i
n
t
u
r
n
i
s
u
s
ed
to
co
m
p
u
te
t
h
e
p
ar
am
eter
s
r
eq
u
ir
ed
f
o
r
m
e
m
b
er
s
h
ip
f
u
n
ctio
n
s
t
h
at
n
ee
d
ed
f
o
r
th
e
FIS
to
r
ig
h
tl
y
co
r
r
elate
in
p
u
t o
r
o
u
tp
u
t d
ata
in
f
u
zz
y
d
o
m
ain
.
7
.
5
.
Co
m
pa
ra
t
iv
e
Ana
ly
s
is
T
h
e
co
m
p
ar
ati
v
e
s
t
u
d
y
d
o
n
e
b
y
Mu
r
t
h
y
et
al.
r
ep
o
r
ted
th
a
t
9
2
%
s
u
cc
e
s
s
r
ate
w
a
s
ac
h
ie
v
ed
u
s
in
g
HM
M
w
ith
w
a
v
elet
f
ea
t
u
r
es
[
3
4
]
an
d
estab
lis
h
ed
th
at
t
h
i
s
ap
p
r
o
ac
h
p
er
f
o
r
m
s
m
o
r
e
e
f
f
ici
en
tl
y
in
co
m
p
ar
i
s
o
n
to
SVM
w
it
h
co
lo
r
f
ea
tu
r
e
s
[
3
6
]
an
d
SVM
w
it
h
w
a
v
elet
MR
A
f
ea
t
u
r
es
[
9
]
.
T
h
e
class
i
f
icatio
n
ac
cu
r
ac
ie
s
r
ep
o
r
ted
b
y
r
ec
en
t
liter
at
u
r
e
a
r
e
p
r
esen
ted
i
n
t
h
e
T
ab
le
5
.
D
u
e
to
i
ts
ac
c
u
r
ac
y
o
f
i
n
s
u
lato
r
d
a
m
a
g
e
d
etec
tio
n
,
th
e
HM
M
lead
s
to
q
u
ick
er
m
ain
te
n
an
ce
a
n
d
r
esto
r
atio
n
o
f
p
o
w
er
s
u
p
p
l
y
co
m
p
ar
ed
to
t
h
e
u
s
e
o
f
SVM
[
3
6
,
3
7
]
.
I
t
is
also
s
h
o
w
n
th
a
t
FI
S,
w
it
h
it
s
i
n
h
er
e
n
t
ab
ilit
y
,
c
o
m
p
u
tes
t
h
e
h
ea
lth
o
f
in
s
u
la
t
o
r
s
in
ter
m
s
o
f
t
h
e
d
eg
r
ee
to
w
h
ic
h
th
e
i
n
s
u
lato
r
s
ar
e
in
g
o
o
d
,
m
ar
g
i
n
al
o
r
r
is
k
y
s
tate
s
.
A
co
m
p
ar
is
o
n
b
etw
ee
n
th
e
m
eth
o
d
s
,
DOST
-
A
NFI
S
a
n
d
DOST
-
SV
M
w
as
d
o
n
e
b
y
R
e
d
d
y
et
al.
[
1
]
in
co
r
p
o
r
atin
g
t
h
e
co
m
p
lex
b
ac
k
g
r
o
u
n
d
i
m
a
g
es
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
201
7
:
3
0
3
–
3
1
2
310
an
d
r
es
u
lts
p
r
o
v
e
t
h
e
e
f
f
ec
ti
v
en
e
s
s
o
f
th
e
p
r
o
p
o
s
ed
tech
n
iq
u
e
s
i
n
d
ea
li
n
g
w
it
h
d
i
f
f
er
en
t
p
o
s
s
ib
ilit
ies
o
f
co
m
p
le
x
b
ac
k
g
r
o
u
n
d
s
.
T
h
e
y
w
o
u
ld
o
th
er
w
is
e
lead
to
w
r
o
n
g
co
n
cl
u
s
io
n
s
ab
o
u
t t
h
e
co
n
d
it
io
n
o
f
in
s
u
lato
r
s
.
T
ab
le
5
C
o
m
p
ar
is
o
n
o
f
Feat
u
r
e
E
x
tr
ac
tio
n
an
d
De
f
ec
t C
lass
i
f
icatio
n
T
ec
h
n
iq
u
e
s
C
l
a
ssi
f
i
c
a
t
i
o
n
A
p
p
r
o
a
c
h
F
e
a
t
u
r
e
v
e
c
t
o
r
t
y
p
e
C
l
a
ssi
f
i
c
a
t
i
o
n
A
c
c
u
r
a
c
y
(
%)
R
e
f
e
r
e
n
c
e
N
o
.
S
V
M
C
o
l
o
u
r
85
[
3
6
]
S
V
M
W
a
v
e
l
e
t
90
[
9
]
H
M
M
W
a
v
e
l
e
t
92
[
3
4
]
A
N
F
I
S
D
O
S
T
-
[
3
5
]
S
V
M
&
A
N
F
I
S
D
O
S
T
-
[
1
]
8.
SCO
P
E
F
O
R
F
UT
UR
E
WO
RK
T
h
er
e
ar
e
s
ev
er
al
to
o
ls
b
o
th
i
n
t
h
e
s
p
atia
l
a
n
d
f
r
eq
u
e
n
c
y
d
o
m
a
in
s
li
k
e
lo
ca
l
b
in
ar
y
p
atte
r
n
(
L
B
P
)
,
cu
r
v
elet
tr
a
n
s
f
o
r
m
,
an
d
co
n
to
u
r
let
tr
an
s
f
o
r
m
etc.
,
f
o
r
f
ea
t
u
r
e
ex
tr
ac
tio
n
.
Dee
p
b
elief
n
et
wo
r
k
s
(
DB
Ns)
w
h
ic
h
is
a
r
ep
r
esen
tati
v
e
m
eth
o
d
o
f
d
ee
p
lear
n
in
g
[
7
1
]
an
d
E
x
tr
e
m
e
L
ea
r
n
i
n
g
Ma
ch
i
n
e
(
E
L
M)
ca
n
b
e
u
s
ed
f
o
r
class
i
f
icatio
n
.
T
h
e
n
e
w
l
y
e
m
er
g
ed
m
ac
h
i
n
e
lear
n
i
n
g
th
eo
r
y
,
d
ee
p
lear
n
in
g
m
a
y
b
e
ap
p
lied
f
o
r
b
etter
class
i
f
icatio
n
w
h
ic
h
o
u
tp
er
f
o
r
m
ed
t
h
e
o
th
er
s
tate
-
o
f
-
t
h
e
-
ar
t
class
i
f
icat
io
n
m
et
h
o
d
s
.
T
h
e
v
ar
io
u
s
tech
n
iq
u
es
p
r
o
p
o
s
ed
b
y
th
e
d
i
f
f
er
en
t
a
u
t
h
o
r
s
d
id
n
o
t
co
n
s
id
er
an
y
o
f
t
h
e
n
o
n
-
v
i
s
u
al
d
ef
ec
t
s
s
u
ch
as
i
n
ter
n
al
cr
ac
k
s
,
h
ig
h
p
ar
tial
d
is
ch
ar
g
e
ac
tiv
ities
a
n
d
h
ig
h
lea
k
a
g
e
cu
r
r
en
t.
T
h
er
m
al
ca
m
er
as
w
o
u
ld
b
e
a
v
iab
le
o
p
tio
n
to
tak
e
th
ese
asp
ec
ts
i
n
to
co
n
s
id
er
atio
n
.
T
h
e
g
eo
g
r
ap
h
ical
i
n
f
er
en
ce
s
y
s
te
m
(
GI
S)
m
a
y
b
e
u
s
ed
to
o
b
tain
s
p
atial
co
o
r
d
in
ates f
o
r
d
is
tr
ib
u
tio
n
s
y
s
te
m
p
la
n
n
in
g
a
n
d
au
to
m
atio
n
.
On
ce
t
h
e
i
m
ag
e
s
s
e
n
t b
y
t
h
e
R
T
Us ar
e
o
b
tain
ed
,
th
e
GI
S
-
aid
ed
in
s
u
lato
r
m
o
n
i
to
r
in
g
w
o
u
ld
r
ed
u
ce
th
e
e
x
is
t
in
g
p
r
o
b
le
m
s
b
y
as
s
i
g
n
i
n
g
u
n
iq
u
e
id
en
t
if
icatio
n
n
u
m
b
er
s
to
R
T
Us f
o
r
s
u
b
s
eq
u
en
t i
m
ag
e
p
r
o
ce
s
s
i
n
g
an
d
p
r
o
p
er
m
o
n
ito
r
in
g
o
f
t
h
e
i
n
s
u
lato
r
s
.
RE
F
E
R
E
NC
E
S
[1
]
M
.
Ja
y
a
Bh
a
ra
ta
Re
d
d
y
,
Ka
rth
ik
Ch
a
n
d
ra
B
a
n
d
D
.
K.
M
o
h
a
n
ta,
"
Co
n
d
i
ti
o
n
M
o
n
i
to
ri
n
g
o
f
1
1
k
V
Distrib
u
ti
o
n
S
y
st
e
m
In
su
lato
rs In
c
o
rp
o
ra
ti
n
g
Co
m
p
lex
I
m
a
g
e
r
y
Us
in
g
Co
m
b
in
e
d
DO
S
T
-
S
V
M
A
p
p
ro
a
c
h
"
,
IEE
E
T
ra
n
sa
c
ti
o
n
s o
n
Die
lec
trics
a
n
d
El
e
c
trica
l
I
n
su
l
a
ti
o
n
,
Vo
l.
2
0
,
Iss
u
e
2
p
p
.
6
6
4
-
6
7
4
,
A
p
ril
2
0
1
3
.
[2
]
Da
b
o
Z
h
a
n
g
,
W
e
n
y
u
a
n
L
i,
F
e
ll
o
w
,
IEE
E,
a
n
d
X
iao
f
u
X
i
o
n
g
,
“
Ov
e
rh
e
a
d
L
in
e
P
re
v
e
n
ti
v
e
M
a
in
t
e
n
a
n
c
e
S
trate
g
y
Ba
se
d
o
n
Co
n
d
it
io
n
M
o
n
it
o
ri
n
g
a
n
d
S
y
ste
m
Re
li
a
b
il
it
y
A
s
se
ss
m
e
n
t”,
IEE
E
T
ra
n
s
o
n
Po
we
r
S
y
ste
ms
,
V
o
l.
2
9
,
NO
.
4
,
Ju
ly
2
0
1
4
1
8
3
9
.
[3
]
C.
J.
Kim
,
Je
o
n
g
Ho
o
n
S
h
in
,
M
y
e
o
n
g
-
Ho
Yo
G
iW
o
n
L
e
e
,
“
A
S
tu
d
y
o
n
th
e
Ch
a
ra
c
teriz
a
ti
o
n
o
f
th
e
In
c
ip
ien
t
F
a
il
u
re
Be
h
a
v
io
r
o
f
In
su
lato
rs
i
n
P
o
w
e
r
Distrib
u
ti
o
n
L
in
e
”
,
“
IEE
E
T
ra
n
s
a
c
ti
o
n
s
o
n
Po
we
r
De
li
v
e
ry
”
,
Vo
l
.
1
4
,
No
.
2
,
A
p
ril
1
9
9
9
p
p
5
1
9
-
5
2
4
.
[4
]
R.
K.
Ag
g
a
r
w
a
l,
A
.
T
.
Jo
h
n
s,
J.A.S
.
B.
Ja
y
a
sin
g
h
e
,
“
A
n
o
v
e
r
v
ie
w
o
f
th
e
c
o
n
d
it
io
n
m
o
n
i
to
ri
n
g
o
f
o
v
e
rh
e
a
d
li
n
e
s”
,
El
e
c
tric P
o
we
r S
y
ste
ms
Res
e
a
rc
h
5
3
(2
0
0
0
)
1
5
–
2
2
,
,
W
.
S
u
,
p
p
.
1
5
-
22
[5
]
H.
H.
Ko
rd
k
h
e
il
i,
H.
A
b
ra
v
e
sh
,
M
.
T
a
b
a
si,
M
.
Da
k
h
e
m
,
a
n
d
M
.
M
.
A
b
ra
v
e
sh
,
“
D
e
ter
m
in
in
g
th
e
p
ro
b
a
b
il
it
y
o
f
fl
a
sh
o
v
e
r
o
c
c
u
rre
n
c
e
in
c
o
m
p
o
sit
e
in
su
lato
rs b
y
u
sin
g
le
a
k
a
g
e
c
u
rre
n
t
h
a
rm
o
n
ic co
m
p
o
n
e
n
ts,
”
IEE
E
T
ra
n
s.
Die
lec
tr.
El
e
c
tr.
In
su
l
.
,
v
o
l.
1
7
,
n
o
.
2
,
p
p
.
5
0
2
–
5
1
2
,
2
0
1
0
.
[6
]
C.
A
n
d
re
a
,
e
t
a
l,
“
In
f
e
rrin
g
c
e
ra
m
ic
in
su
lato
r
p
o
l
lu
ti
o
n
b
y
a
n
in
n
o
v
a
ti
v
e
a
p
p
ro
a
c
h
re
so
rti
n
g
to
P
D
d
e
tec
ti
o
n
,
”
IEE
E
T
ra
n
s.
Die
lec
tr.
El
e
c
tr.
I
n
su
l
.
,
v
o
l
.
1
4
,
n
o
.
1
,
p
p
.
2
3
–
2
9
,
2
0
0
7
[7
]
S
.
V
e
n
k
a
tara
m
a
n
,
e
t
a
l,
“
I
m
p
a
c
t
o
f
w
e
a
th
e
rin
g
o
n
fl
a
sh
o
v
e
r
p
e
rf
o
rm
a
n
c
e
o
f
n
o
n
c
e
ra
m
i
c
in
su
lato
r
s,”
I
EE
E
T
ra
n
s.
Die
lec
ti
c
.
El
e
c
tr.
In
su
l.
,
v
o
l.
1
5
,
n
o
.
4
,
p
p
.
1
0
7
3
–
1
0
8
0
,
2
0
0
8
.
[8
]
I.
Ra
m
irez
-
V
a
z
q
u
e
z
,
R.
He
rn
a
n
d
e
z
-
Co
ro
n
a
,
a
n
d
G
.
M
o
n
t
o
y
a
-
Te
n
a
,
“
Dia
g
n
o
stics
f
o
r
n
o
n
c
e
ra
m
i
c
i
n
su
lato
rs
i
n
h
a
rs
h
e
n
v
iro
n
m
e
n
ts,
”
IEE
E
El
e
c
tr.
In
s
u
l.
M
a
g
.
,
v
o
l.
2
5
,
n
o
.
6
,
p
p
.
2
8
–
3
3
,
2
0
0
9
.
[9
]
V
.
S
.
M
u
rt
h
y
,
e
t
a
l,
“
In
su
lato
r
c
o
n
d
it
io
n
a
n
a
ly
sis
f
o
r
o
v
e
rh
e
a
d
d
i
strib
u
ti
o
n
li
n
e
s
u
sin
g
c
o
m
b
in
e
d
w
a
v
e
let
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
,
”
IEE
E
T
ra
n
s.
Di
e
lec
tr
.
El
e
c
tr.
In
su
l
.
,
v
o
l.
1
7
,
n
o
.
1
,
p
p
.
8
9
–
9
9
,
2
0
1
0
.
[1
0
]
M
.
Ko
m
o
d
a
,
e
t
a
l
,
“
El
e
c
tro
m
a
g
n
e
ti
c
in
d
u
c
ti
o
n
m
e
th
o
d
f
o
r
d
e
tec
ti
n
g
a
n
d
l
o
c
a
ti
n
g
fl
a
w
s
o
n
o
v
e
rh
e
a
d
tran
sm
issio
n
li
n
e
s,”
IEE
E
T
ra
n
s
.
Po
we
r De
l.
,
v
o
l.
5
,
n
o
.
3
,
p
p
.
1
4
8
4
–
4
9
0
,
1
9
9
0
.
[1
1
]
H.Z
a
n
g
l
e
t
a
l,
“
A
fe
a
sib
il
it
y
stu
d
y
o
n
a
u
to
n
o
m
o
u
s
o
n
li
n
e
c
o
n
d
it
io
n
m
o
n
it
o
r
in
g
o
f
h
ig
h
-
v
o
lt
a
g
e
o
v
e
rh
e
a
d
p
o
w
e
r
li
n
e
s,”
IEE
E
T
ra
n
s
.
In
str
u
m.
M
e
a
s.,
v
o
l.
5
8
,
n
o
.
5
,
p
p
.
1
7
8
9
–
1
7
9
6
,
2
0
0
9
.
[1
2
]
T
.
Hja
rtars
o
n
,
e
t
a
l,
“
De
v
e
lo
p
m
e
n
t
o
f
h
e
a
lt
h
in
d
ice
s
f
o
r
a
s
se
t
c
o
n
d
it
io
n
a
ss
e
ss
m
e
n
t,
”
in
P
r
o
c
.
IEE
E
T
ra
n
s.
a
n
d
Dist
.
Co
n
f.
Exp
o
.
,
2
0
0
3
,
v
o
l.
2
,
p
p
.
5
4
1
–
44.
[1
3
]
Y.
Ha
n
a
n
d
Y.
H.
S
o
n
g
,
“
Co
n
d
it
i
o
n
M
o
n
it
o
ri
n
g
T
e
c
h
n
iq
u
e
s
f
o
r
El
e
c
tri
c
a
l
Eq
u
ip
m
e
n
t
—
A
L
it
e
ra
tu
re
S
u
rv
e
y
”
,
IEE
E
tra
n
sa
c
ti
o
n
s
o
n
p
o
we
r d
e
li
v
e
ry
,
v
o
l.
1
8
,
n
o
.
1
,
p
p
.
4
-
1
3
,
Ja
n
.
2
0
0
3
[1
4
]
D.
I.
Jo
n
e
s,
C.
C.
W
h
it
w
o
rth
,
G
.
K.
Earp
a
n
d
A
.
W
.
G
.
Du
ll
e
r,
A
lab
o
ra
t
o
ry
tes
t
-
b
e
d
f
o
r
a
n
a
u
t
o
m
a
ted
p
o
w
e
r
li
n
e
in
sp
e
c
ti
o
n
sy
ste
m
,
Co
n
tro
l
E
n
g
i
n
e
e
rin
g
Pra
c
ti
c
e
,
V
o
l.
1
3
,
No
.
7
,
p
p
.
8
3
5
–
8
5
1
,
2
0
0
5
.
[1
5
]
D.
I.
J
o
n
e
s
a
n
d
G
.
K.
Earp
,
Ca
m
e
ra
sig
h
tl
in
e
p
o
i
n
ti
n
g
re
q
u
ire
m
e
n
ts
f
o
r
a
e
rial
in
sp
e
c
ti
o
n
o
f
o
v
e
rh
e
a
d
p
o
w
e
r
li
n
e
s,
El
e
c
tric P
o
we
r S
y
ste
ms
Res
e
a
rc
h
,
Vo
l.
5
7
,
p
p
.
7
3
–
8
2
,
2
0
0
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
R
ev
iew
o
f Ma
ch
in
e
V
is
io
n
b
a
s
ed
I
n
s
u
la
to
r
I
n
s
p
ec
tio
n
S
yste
ms fo
r
Ove
r
h
ea
d
.
..
(
P
.
S
u
r
ya
P
r
a
s
a
d
)
311
[1
6
]
C.
C.
W
h
it
w
o
rth
,
e
t
a
l,
“
A
e
ri
a
l
v
id
e
o
in
sp
e
c
ti
o
n
o
f
p
o
w
e
r
li
n
e
s”
,
Po
we
r
En
g
in
e
e
rin
g
J
.
,
Vo
l.
1
5
,
N
o
.
1
,
p
p
.
2
5
–
3
2
,
2
0
0
1
.
[1
7
]
I
a
n
G
o
li
g
h
tl
y
a
n
d
De
w
i
Jo
n
e
s,
“
Co
rn
e
r
d
e
tec
ti
o
n
a
n
d
m
a
tch
in
g
f
o
r
v
isu
a
l
trac
k
in
g
d
u
rin
g
p
o
w
e
r
li
n
e
i
n
sp
e
c
ti
o
n
,
”
Ima
g
e
a
n
d
Vi
sio
n
C
o
mp
u
ti
n
g
,
v
o
l
.
2
1
,
p
p
.
8
2
7
–
8
4
0
,
2
0
0
3
.
[1
8
]
A
.
W
.
G
Du
ll
e
r
C.
C.
W
h
it
w
o
rth
,
D.
I.
Jo
n
e
s an
d
G
.
K.
Earp
,
“
A
e
ri
a
l
v
id
e
o
in
s
p
e
c
ti
o
n
o
f
o
v
e
rh
e
a
d
p
o
w
e
r
li
n
e
s,”
IE
E
Po
we
r E
n
g
in
e
e
rin
g
J
o
u
rn
a
l
,
v
o
l.
1
5
,
n
o
.
1
,
p
p
.
2
5
–
3
2
,
2
0
0
1
.
[1
9
]
A
e
ro
sp
e
c
t,
“
Uta
h
sta
te
to
re
v
o
lu
ti
o
n
ize
p
o
w
e
r
li
n
e
in
sp
e
c
ti
o
n
s,”
h
tt
p
:/
/w
ww
.
sp
a
c
e
d
a
il
y
.
c
o
m
/n
e
w
s/e
n
e
r
g
y
-
te
c
h
-
0
3
z
e
.
h
tm
l
,
2
0
0
3
.
[2
0
]
D.
I.
Jo
n
e
s,
“
A
e
rial
in
sp
e
c
ti
o
n
o
f
o
v
e
rh
e
a
d
p
o
w
e
r
li
n
e
s
u
sin
g
v
id
e
o
:
Esti
m
a
ti
o
n
o
f
i
m
a
g
e
b
lu
rrin
g
d
u
e
to
v
e
h
icle
a
n
d
c
a
m
e
ra
m
o
ti
o
n
,
”
in
Pro
c
IEE
Vi
si
o
n
,
Ima
g
e
a
n
d
S
ig
n
a
l
Pro
c
e
ss
in
g
,
2
0
0
0
,
p
p
.
1
5
7
–
1
6
6
.
[2
1
]
F
.
Xu
,
e
t
a
l,
“
P
e
d
e
str
ian
d
e
tec
ti
o
n
a
n
d
trac
k
in
g
w
it
h
n
ig
h
t
v
isio
n
”
,
I
EE
E
T
ra
n
s.
I
n
telli
g
e
n
t
T
ra
n
sp
o
rt
a
ti
o
n
S
y
st.
,
V
o
l
.
6
,
p
p
.
6
3
-
7
1
,
2
0
0
5
.
[2
2
]
R.
P
.
G
u
p
ta,
e
t
a
l.
,
“
A
u
to
m
a
te
d
V
e
rse
s
Co
n
v
e
n
ti
o
n
a
l
Distri
b
u
ti
o
n
S
y
ste
m
”
,
Pro
c
.
o
f
th
e
T
h
i
rd
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Po
we
r
a
n
d
E
n
e
rg
y
S
y
ste
ms
Eu
ro
PE
S
-
2
0
0
3
,
S
p
a
i
n
,
2
0
0
3
,
p
p
.
3
3
-
3
8
.
[2
3
]
A
.
P
a
h
w
a
a
n
d
J.
K.
S
h
u
lt
is,
A
ss
e
s
s
m
e
n
t
o
f
th
e
P
re
se
n
t
S
tatu
s o
f
Dis
tri
b
u
t
io
n
A
u
to
m
a
ti
o
n
,
En
g
g
.
Ex
p
e
rime
n
t
S
tatio
n
,
Ka
n
sa
s S
tate
n
iv
.
,
M
a
n
h
a
tt
a
n
,
KS
,
Re
p
.
2
3
8
,
1
2
.
[2
4
]
R.
P
.
G
u
p
ta
a
n
d
R.
K.
V
e
rm
a
,
P
o
w
e
r
S
y
ste
m
A
u
to
m
a
ti
o
n
:
Aca
d
e
mic
o
p
e
n
I
n
ter
n
e
t
jo
u
rn
a
l,
V
o
lu
m
e
1
5
,
2
0
0
5
,
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
a
n
d
Co
m
p
u
ter E
n
g
in
e
e
rin
g
Un
iv
e
rsity
o
f
W
e
ste
rn
On
tario
.
[2
5
]
L
il
i
M
a
,
Ya
n
g
Qu
a
n
Ch
e
n
:
A
e
rial
S
u
rv
e
il
lan
c
e
S
y
st
e
m
f
o
r
o
v
e
rh
e
a
d
p
o
w
e
r
li
n
e
in
sp
e
c
ti
o
n
,
T
e
c
h
n
ica
l
re
p
o
rt
USUCS
OIS
-
TR
-
04
-
0
8
(2
0
0
4
),
Ut
a
h
S
tate
Un
iv
e
rsity
,
USA
.
[2
6
]
B.
A
v
id
a
r:
El
e
c
tro
n
ic
a
irb
o
rn
e
in
sp
e
c
ti
o
n
m
e
th
o
d
s
f
o
r
o
v
e
rh
e
a
d
tran
sm
issio
n
p
o
w
e
r
-
li
n
e
s.
6
t
h
In
t.
Co
n
f.
o
n
T
ra
n
sm
issio
n
a
n
d
Distrib
u
ti
o
n
C
o
n
stru
c
ti
o
n
a
n
d
L
ive
L
in
e
M
a
i
n
te
n
a
n
c
e
,
p
.
8
9
–
9
3
,
1
9
9
3
.
[2
7
]
J.
Ka
tras
n
ik
,
e
t
a
l
“
A
su
rv
e
y
o
f
m
o
b
il
e
ro
b
o
ts
f
o
r
d
istr
ib
u
t
io
n
p
o
w
e
r
li
n
e
in
sp
e
c
ti
o
n
”
,
IEE
E
T
ra
n
s
a
c
ti
o
n
s
o
n
Po
we
r
De
li
v
e
ry
,
2
5
(1
),
2
0
1
0
.
[2
8
]
S
a
m
p
e
d
ro
,
Ca
rlo
s,
e
t
a
l.
"
A
su
p
e
rv
ise
d
a
p
p
ro
a
c
h
to
e
lec
tri
c
to
w
e
r
d
e
tec
ti
o
n
a
n
d
c
las
sif
ica
ti
o
n
f
o
r
p
o
w
e
r
li
n
e
in
sp
e
c
ti
o
n
.
"
Ne
u
ra
l
Ne
tw
o
rk
s (
IJ
CNN),
2
0
1
4
In
ter
n
a
ti
o
n
a
l
J
o
i
n
t
C
o
n
fer
e
n
c
e
o
n
.
IEE
E,
2
0
1
4
.
[2
9
]
P
a
n
,
L
if
e
n
g
,
"
In
telli
g
e
n
t
I
m
a
g
e
Re
c
o
g
n
it
io
n
Re
se
a
rc
h
o
n
S
tat
u
s
o
f
P
o
w
e
r
T
r
a
n
s
m
issio
n
L
in
e
s,"
S
e
n
so
rs
&
T
ra
n
sd
u
c
e
rs
(
1
7
2
6
-
5
4
7
9
)
1
7
9
.
9
2
0
1
4
.
[3
0
]
R.
C.
G
o
n
z
a
lez
,
R.
E.
W
o
o
d
s,
Di
g
it
a
l
Im
a
g
e
Pro
c
e
ss
in
g
,
3
r
d
e
d
n
.
(P
e
a
rso
n
Ed
u
c
a
ti
o
n
,
2
0
0
8
).
IS
BN
9
7
8
-
81
-
3
1
7
-
1
9
3
4
-
3.
[3
1
]
S
.
Ha
y
k
in
s,
Ne
u
ra
l
Ne
two
rk
s
,
2
n
d
e
d
n
.
(
P
e
a
rso
n
Ed
u
c
a
ti
o
n
,
1
9
9
9
).
IS
BN 8
1
-
7
8
0
8
--
3
0
0
-
0
[3
2
]
In
v
e
stig
a
ti
o
n
o
f
a
p
p
ly
in
g
n
e
w
tec
h
n
o
l
o
g
ies
to
o
v
e
rh
e
a
d
tran
sm
i
ss
io
n
li
n
e
i
n
sp
e
c
ti
o
n
s,
P
r
o
jec
t
1
4
9
7
–
2
,
El
e
c
tric
Po
we
r R
e
se
a
rc
h
In
stit
u
te R
e
p
o
rt
,
S
e
p
t
e
m
b
e
r
1
9
8
1
.
[3
3
]
Bo
W
e
n
W
a
n
g
,
Qu
a
n
G
u
,
A
D
e
tec
ti
o
n
M
e
th
o
d
f
o
r
T
ra
n
sm
i
ss
i
o
n
L
in
e
In
su
lat
o
rs
Ba
se
d
o
n
a
n
Im
p
ro
v
e
d
F
CM
A
l
g
o
rit
h
m
,
T
EL
KOM
NIKA
,
V
o
l.
1
3
,
No
.
1
,
M
a
rc
h
2
0
1
5
,
p
p
.
1
6
4
~
1
7
2
[3
4
]
V
e
lag
a
S
re
e
ra
m
a
M
u
rth
y
,
e
t
a
l,
“
Dig
it
a
l
Im
a
g
e
P
ro
c
e
ss
in
g
a
p
p
r
o
a
c
h
u
si
n
g
c
o
m
b
in
e
d
W
a
v
e
let
-
Hid
d
e
n
M
a
rk
o
v
M
o
d
e
l
(HM
M
)
f
o
r
w
e
ll
-
b
e
in
g
a
n
a
l
y
sis
o
f
in
su
lato
rs”
,
In
ter
n
a
t
io
n
a
l
J
o
u
r
n
a
l
o
f
IE
T
ima
g
e
p
ro
c
e
ss
in
g
,
Vo
l.
5
,
Iss
.
2
,
2
0
1
1
,
p
p
1
7
1
-
1
8
3
[3
5
]
M
.
Ja
y
a
Bh
a
ra
ta
Re
d
d
y
,
B.
Ka
rth
ik
Ch
a
n
d
ra
,
D.
K.
M
o
h
a
n
ta,
“
A
DO
S
T
Ba
se
d
A
p
p
ro
a
c
h
f
o
r
th
e
C
o
n
d
it
i
o
n
M
o
n
i
to
r
i
n
g
o
f
1
1
k
V
Distri
b
u
ti
o
n
L
in
e
In
s
u
lato
rs”
,
I
EE
E
T
ra
n
s.
o
n
Die
lec
trics
a
n
d
El
e
c
trica
l
I
n
s
u
la
ti
o
n
,
V
o
l.
1
8
,
Iss
u
e
2
,
p
p
.
5
8
8
-
5
9
5
,
A
p
ril
2
0
1
1
.
[3
6
]
V
.
S
.
M
u
rt
h
y
,
e
t
a
l
,
Distri
b
u
ti
o
n
sy
ste
m
in
su
lato
r
m
o
n
it
o
rin
g
u
sin
g
v
id
e
o
s
u
rv
e
il
lan
c
e
a
n
d
S
VM
f
o
r
c
o
m
p
lex
b
a
c
k
g
ro
u
n
d
im
a
g
e
s,
In
t.
J
o
u
rn
a
l
o
f
P
o
we
r a
n
d
e
n
e
rg
y
c
o
n
v
e
rs
io
n
,
V
o
l
.
1
,
No
.
1
,
p
p
4
9
-
7
2
,
2
0
0
9
[3
7
]
S
.
M
u
rth
y
,
D.
M
o
h
a
n
ta,
S
.
G
u
p
t
a
,
Distrib
u
ti
o
n
sy
ste
m
in
su
lato
r
m
o
n
it
o
rin
g
u
si
n
g
v
id
e
o
s
u
rv
e
il
lan
c
e
a
n
d
S
VM
f
o
r
c
o
m
p
lex
b
a
c
k
g
ro
u
n
d
im
a
g
e
s,
In
t.
J
n
l.
o
f
Co
m
p
Ap
p
l
a
n
d
T
e
c
h
.
,
V
o
l
3
,
N
o
.
1
,
p
p
1
1
-
3
1
,
2
0
1
1
[3
8
]
Zh
u
,
Hu
,
W
.
G
.
L
i,
a
n
d
Ye
L
in
.
"
P
re
se
n
t
a
n
d
f
u
tu
re
d
e
v
e
lo
p
m
e
n
t
o
f
d
e
tec
ti
o
n
m
e
th
o
d
s
f
o
r
c
o
m
p
o
site
in
su
lato
r.
"
In
su
l
a
t
o
rs
a
n
d
S
u
rg
e
Arre
ste
rs
8
.
1
(2
0
0
6
):
1
3
3
-
1
3
7
.
[3
9
]
T
G
u
o
,
W
W
a
n
g
,
H
Ya
n
g
,
X
Yu
a
n
”
Re
se
a
rc
h
o
n
li
v
e
li
n
e
a
u
to
-
tes
ti
n
g
tec
h
n
o
lo
g
y
f
o
r
tran
sm
issio
n
li
n
e
in
s
u
lato
rs."
(2
0
1
5
),
3
rd
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
M
e
c
h
a
tr
o
n
ics
,
Ro
b
o
ti
c
s
a
n
d
Au
to
ma
ti
o
n
.
[4
0
]
Qa
d
d
o
u
m
i,
e
t
a
l,
"
Ou
td
o
o
r
I
n
s
u
lato
rs
T
e
stin
g
Us
in
g
A
rti
f
icia
l
Ne
u
ra
l
Ne
t
w
o
rk
-
Ba
s
e
d
Ne
a
r
-
F
ield
M
icro
w
a
v
e
T
e
c
h
n
iq
u
e
.
"
In
stru
me
n
t
a
ti
o
n
a
n
d
M
e
a
su
re
me
n
t,
IEE
E
T
r
a
n
s
o
n
6
3
.
2
(2
0
1
4
):
2
6
0
-
2
6
6
.
[4
1
]
P
a
l
o
AL
to
,
“
EP
RI
a
e
rial
in
sp
e
c
ti
o
n
sy
ste
m
a
ll
se
t
f
o
r
tak
e
o
ff
,
”
h
tt
p
:/
/w
ww
.
e
p
ri.
c
o
m
/
c
o
rp
o
ra
te/d
isc
o
v
e
r
,
e
p
ri/
n
e
w
s/
2
0
0
1
re
lea
se
s/ 0
1
1
0
1
7
a
e
rial.
h
tm
l,
2
0
0
1
.
[4
2
]
R.
P
.
G
u
p
ta,
a
n
d
S
.
C
.
S
riv
a
sta
v
a
,
“
T
e
c
h
n
o
lo
g
y
d
e
v
e
lo
p
m
e
n
t
a
n
d
im
p
lem
e
n
tatio
n
f
o
r
p
o
w
e
r
d
istri
b
u
ti
o
n
a
u
to
m
a
ti
o
n
”
,
W
a
ter
a
n
d
En
e
rg
y
I
n
ter
n
a
ti
o
n
a
l
J
.
,
V
o
l
.
6
1
,
p
p
.
4
0
-
4
7
,
2
0
0
4
.
[4
3
]
Du
tt
a
,
T
a
n
im
a
,
e
t
a
l.
"
I
m
a
g
e
A
n
a
l
y
sis
-
Ba
se
d
A
u
to
m
a
ti
c
D
e
tec
ti
o
n
of
T
ra
n
s
m
issio
n
T
o
we
rs
u
sin
g
A
e
ri
a
l
Im
a
g
e
r
y
.
"
Pa
tt
e
rn
Rec
o
g
n
it
io
n
a
n
d
Im
a
g
e
A
n
a
lys
is.
S
p
ri
n
g
e
r
In
tern
a
ti
o
n
a
l
P
u
b
l
ish
i
n
g
,
2
0
1
5
.
6
4
1
-
6
5
1
.
[4
4
]
M
a
rti
n
e
z
e
t
a
l,
T
o
w
a
rd
s
a
u
to
n
o
m
o
u
s
a
n
d
trac
k
in
g
o
f
e
le
c
tri
c
to
w
e
r
s
f
o
r
a
e
rial
p
o
we
r
li
n
e
in
sp
e
c
ti
o
n
,
Un
m
a
n
n
e
d
Ai
rc
ra
ft
S
y
ste
ms
(
ICUAS
),
2
0
1
4
I
n
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
[4
5
]
A
.
P
h
i
li
p
s
e
t
a
l:
A
irb
o
r
n
e
In
s
p
e
c
ti
o
n
T
e
c
h
n
o
lo
g
y
:
M
a
rk
e
t
S
u
rv
e
y
.
T
e
c
h
n
ica
l
Re
p
o
rt
1
0
0
6
7
4
9
,
EP
RI
,
P
a
l
o
A
lt
o
,
Ca
li
f
o
rn
ia ,
USA
2
0
0
2
.
[4
6
]
M
a
z
u
re
k
,
e
t
a
l.
"
A
p
p
li
c
a
ti
o
n
o
f
b
a
c
k
g
ro
u
n
d
e
stim
a
ti
o
n
a
n
d
re
m
o
v
a
l
tec
h
n
iq
u
e
s
f
o
r
th
e
e
x
trac
ti
o
n
o
f
th
e
p
o
w
e
r
li
n
e
c
o
m
p
o
n
e
n
ts
o
n
th
e
d
ig
it
a
l
im
a
g
e
s
f
o
r
th
e
a
u
to
m
a
ti
c
p
o
w
e
r
li
n
e
i
n
s
p
e
c
ti
o
n
sy
ste
m
s."
P
o
m
iar
y
,
A
u
to
m
a
t
y
k
a
,
Ko
n
tro
la
5
4
(
2
0
0
8
):
6
9
8
-
6
9
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
201
7
:
3
0
3
–
3
1
2
312
[4
7
]
L
a
h
iri
,
e
t
a
l.
"
Im
p
o
rtan
c
e
o
f
d
istri
b
u
ti
o
n
a
u
t
o
m
a
ti
o
n
sy
ste
m
f
o
r
In
d
ian
p
o
w
e
r
u
ti
li
ty
.
"
P
o
we
r
&
En
e
rg
y
S
o
c
.
Ge
n
.
M
e
e
ti
n
g
2
0
0
9
PE
S
'0
9
IEE
E.
[4
8
]
D.
Ba
ss
e
tt
,
K.
Cli
n
a
rd
,
J.
G
r
a
in
g
e
r,
S
.
P
u
ru
c
k
e
r,
a
n
d
D.
W
a
rd
,
“
T
u
to
rial
Co
u
rse
:
Distrib
u
ti
o
n
A
u
to
m
a
ti
o
n
”
,
IEE
E
T
u
to
ri
a
l
Pu
b
li
c
a
ti
o
n
8
8
EH0
2
8
0
-
8
-
PW
R
,
1
9
8
8
.
[4
9
]
J.
B.
Bu
n
c
h
,
“
G
u
id
e
li
n
e
s f
o
r
Ev
a
lu
a
ti
n
g
Distrib
u
ti
o
n
A
u
to
m
a
ti
o
n
”
,
EP
RI
Re
p
o
r
t
EL
-
3
7
2
8
,
1
9
8
4
.
[5
0
]
S
.
S
.
V
e
n
k
a
ta,
e
t
a
l.
,
“
W
h
a
t
f
u
tu
re
d
istri
b
u
ti
o
n
e
n
g
in
e
e
rs
n
e
e
d
to
l
e
a
rn
”
,
IEE
E
T
ra
n
s.
Po
we
r
S
y
st
.
,
V
o
l
.
1
9
,
p
p
.
1
7
-
2
3
,
2
0
0
4
.
[5
1
]
Ch
a
n
d
ra
se
k
a
r,
S
.
,
e
t
a
l.
"
In
v
e
stig
a
ti
o
n
s
o
n
lea
k
a
g
e
c
u
rre
n
t
a
n
d
p
h
a
s
e
a
n
g
le
c
h
a
ra
c
teristics
o
f
p
o
rc
e
lain
a
n
d
p
o
ly
m
e
ri
c
in
su
lato
r
u
n
d
e
r
c
o
n
tam
in
a
ted
c
o
n
d
it
io
n
s."
Die
lec
trics
a
n
d
El
e
c
trica
l
In
su
l
a
ti
o
n
,
IEE
E
T
ra
n
s
a
c
ti
o
n
s
o
n
1
6
.
2
(2
0
0
9
)
:
574
-
5
8
3
.
[5
2
]
J.S
.
T
.
L
o
o
m
s,
In
su
lato
rs f
o
r
Hig
h
Vo
lt
a
g
e
s,
P
e
ter
P
e
re
g
rin
u
s L
td
,
1
9
8
8
.
[5
3
]
E.
A
.
Ch
e
rn
e
y
,
Ce
m
e
n
t
g
ro
w
th
f
a
il
u
re
o
f
p
o
rc
e
lain
su
sp
e
n
si
o
n
i
n
su
lato
rs,
IEE
E
T
ra
n
s.
P
AS
-
1
0
2
(8
)
(1
9
8
3
)
2
7
6
5
–
2
7
7
4
.
[5
4
]
E.
A
.
Ch
e
rn
e
y
,
Ce
m
e
n
t
g
ro
w
th
fa
il
u
re
m
e
c
h
a
n
ism
in
p
o
rc
e
lain
s
u
sp
e
n
sio
n
i
n
su
lat
o
rs,
IEE
E
T
ra
n
s.
PW
RD
-
2
(1
)
(1
9
8
7
)
2
4
9
–
2
5
5
.
[5
5
]
A
.
L
.
Ba
rc
la
y
,
D.
A
.
S
w
if
t,
C
a
p
a
n
d
p
i
n
in
su
lato
rs:
e
lec
tri
c
a
l
p
u
n
c
tu
re
o
f
p
o
rc
e
lain
u
n
d
e
r
A
C
e
n
e
rg
isa
ti
o
n
,
5
th
In
ter
n
a
t
io
n
a
l
C
o
n
f
o
n
Die
lec
tric M
a
ter
ia
ls
,
M
e
a
su
re
m
e
n
ts
a
n
d
A
p
p
li
c
a
ti
o
n
s,
1
9
8
8
,
p
p
.
3
7
0
–
3
7
4
[5
6
]
R.
G
.
S
to
c
k
w
e
ll
,
“
A
b
a
sis
f
o
r
e
ff
icie
n
t
re
p
re
se
n
tatio
n
o
f
th
e
S
-
tran
sf
o
r
m
”
,
J.
Dig
it
a
l
S
ig
n
a
l
P
ro
c
e
ss
in
g
,
V
o
l.
1
7
,
p
p
371
–
3
9
3
,
2
0
0
7
.
[5
7
]
S
.
Dra
b
y
c
z
,
R.
G
.
S
to
c
k
w
e
ll
a
n
d
J.
R.
M
it
c
h
e
ll
,
“
Im
a
g
e
te
x
tu
re
c
h
a
ra
c
teriz
a
ti
o
n
u
sin
g
th
e
d
isc
r
e
te
o
rth
o
g
o
n
a
l
s
-
tran
sf
o
r
m
”
,
J
.
Dig
it
a
l
Ima
g
in
g
,
Vo
l.
2
2
,
p
p
.
6
9
6
-
7
0
8
,
2
0
0
9
.
[5
8
]
Y.
W
a
n
g
a
n
d
J.
Or
c
h
a
rd
,
“
T
h
e
d
isc
re
te
o
rth
o
n
o
rm
a
l
S
to
c
k
w
e
ll
tra
n
sf
o
rm
f
o
r
i
m
a
g
e
re
sto
ra
ti
o
n
”
,
1
6
th
IEE
E
In
ter
n
.
Co
n
f.
Ima
g
e
Pro
c
e
ss
in
g
(
ICIP
),
p
p
.
2
7
6
1
-
2
7
6
4
,
(DO
I:
1
0
.
1
1
0
9
/ICIP
.
2
0
0
9
.
5
4
1
4
1
3
5
)
,
2
0
0
9
.
[5
9
]
L
.
R.
Ra
b
in
e
r,
“
A
T
u
to
rial
o
n
Hid
d
e
n
M
a
rk
o
v
M
o
d
e
ls
a
n
d
S
e
lec
ted
A
p
p
li
c
a
ti
o
n
s
in
S
p
e
e
c
h
Re
c
o
g
n
it
i
o
n
”
,
Pro
c
e
e
d
in
g
s
o
f
t
h
e
IEE
E
,
V
o
l
.
7
7
,
No
.
2
,
p
p
.
2
6
7
-
2
9
5
,
1
9
8
9
.
[6
0
]
S. E
ick
e
ler,
e
t
a
l,
”
Hid
d
e
n
M
a
rk
o
v
M
o
d
e
l
Ba
se
d
Co
n
ti
n
u
o
u
s On
li
n
e
Ge
stu
re
Re
c
o
g
n
it
io
n
”
,
14
th
I
n
t.
Co
n
f.
o
n
Pa
t
ter
n
Rec
o
g
n
it
io
n
,
V
o
l.
2
,
p
p
.
1
2
0
6
-
1
2
0
8
,
1
9
9
8
.
[6
1
]
Ja
y
a
sh
re
e
P
a
d
m
a
n
a
b
h
a
n
,
M
e
lv
in
Jo
se
Jo
h
n
so
n
P
re
m
k
u
m
a
r,
M
a
c
h
in
e
L
e
a
rn
in
g
in
A
u
to
m
a
ti
c
S
p
e
e
c
h
Re
c
o
g
n
it
io
n
:
A
S
u
rv
e
y
,
IET
E
T
e
c
h
n
ica
l
Rev
iew
,
V
o
l
.
3
2
,
Iss
.
4,
p
p
.
2
4
0
-
2
5
1
,
Ju
ly
-
A
u
g
.
2
0
1
5
[6
2
]
A
.
V
.
Ne
f
ian
a
n
d
M
.
H.
Ha
y
e
s,
“
F
a
c
e
De
te
c
ti
o
n
a
n
d
Re
c
o
g
n
it
io
n
Us
in
g
Hid
d
e
n
M
a
rk
o
v
M
o
d
e
ls”
,
In
t.
C
o
n
f.
o
n
Ima
g
e
Pro
c
e
ss
in
g
,
Ch
ica
g
o
,
Vo
l.
1
,
p
p
.
1
4
1
-
1
4
5
,
1
9
9
8
.
[6
3
]
M
u
h
a
m
m
a
d
Ha
m
e
e
d
S
id
d
iq
i,
e
t
a
l.
De
p
th
Ca
m
e
ra
-
B
a
se
d
F
a
c
i
a
l
Ex
p
re
ss
io
n
Re
c
o
g
n
it
io
n
S
y
ste
m
Us
in
g
M
u
lt
il
a
y
e
r
S
c
h
e
m
e
,
IET
E
T
e
c
h
n
ica
l
Rev
iew
,
V
o
l
.
3
1
,
Iss
.
4
,
p
p
.
2
7
7
-
2
8
6
,
S
e
p
t
-
Oc
t.
2
0
1
5
.
[6
4
]
C.
Bu
rg
e
s,
“
A
T
u
to
rial
o
n
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
s
f
o
r
P
a
tt
e
rn
R
e
c
o
g
n
it
io
n
”
In
U.
F
a
y
y
a
d
,
e
d
it
o
r
,
Pro
c
e
e
d
in
g
s
o
f
Da
ta
M
in
in
g
a
n
d
Kn
o
wled
g
e
Dis
c
o
v
e
ry
,
p
p
.
1
–
4
3
,
1
9
9
8
.
[6
5
]
Yo
n
g
W
a
n
g
,
Hu
i
G
u
o
,
“
S
h
a
p
e
Re
c
o
g
n
it
io
n
o
f
Ty
re
M
a
rk
in
g
P
o
in
ts
Ba
se
d
o
n
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
”
,
IET
E
T
e
c
h
n
ica
l
Rev
iew
,
Vo
l.
3
2
,
Iss
.
2
,
p
p
.
1
2
3
-
1
3
0
,
M
a
r
-
A
p
r.
2
0
1
5
.
[6
6
]
T
.
Ka
n
u
n
g
o
,
D.M
.
M
o
u
n
t
,
N.
S
.
Ne
tan
y
a
h
u
,
C.
D.
P
iat
k
o
,
R.
S
il
v
e
r
m
a
n
a
n
d
A
.
Y.
W
u
,
"
A
n
e
ff
icie
n
t
k
-
m
e
a
n
s
c
lu
ste
rin
g
a
lg
o
rit
h
m
:
A
n
a
l
y
si
s
a
n
d
im
p
le
m
e
n
tatio
n
"
,
IEE
E
T
ra
n
s.
Pa
tt
e
rn
A
n
a
lys
is
a
n
d
M
a
c
h
i
n
e
In
telli
g
e
n
c
e
,
V
o
l
.
2
4
,
p
p
8
8
1
–
8
9
2
,
2
0
0
2
.
[6
7
]
Co
rtes
,
C.
a
n
d
V
a
p
n
ik
,
V
.
(
1
9
9
5
)
‘S
u
p
p
o
rt
v
e
c
to
r
n
e
tw
o
rk
s’,
M
a
c
h
i
n
e
L
e
a
rn
i
n
g
,
V
o
l
.
2
0
,
p
p
.
1
–
2
5
.
[6
8
]
S
.
S
.
Ke
e
rth
i
a
n
d
C.
J.
L
in
.
“
As
y
m
p
to
ti
c
b
e
h
a
v
io
rs
o
f
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
s
w
it
h
Ga
u
ss
ian
k
e
rn
e
l”,
Ne
u
ra
l
Co
mp
u
t
a
ti
o
n
,
V
o
l
.
1
5
,
p
p
.
1
6
6
7
-
1
6
8
9
,
2
0
0
3
.
[6
9
]
S
.
G
.
M
a
ll
a
t,
“
A
T
h
e
o
r
y
f
o
r
m
u
lt
i
re
so
l
u
ti
o
n
s
ig
n
a
l
d
e
c
o
m
p
o
si
ti
o
n
:
T
h
e
wa
v
e
let
r
e
p
re
se
n
tatio
n
”
,
IEE
E
T
ra
n
s.
Pa
tt
e
rn
A
n
a
lys
is
M
a
c
h
in
e
I
n
telli
g
e
n
c
e
,
V
o
l.
1
1
,
p
p
.
6
7
4
-
6
9
3
,
1
9
8
9
.
[7
0
]
M
.
J.B.
Re
d
d
y
,
D.K.
M
o
h
a
n
ta,
“
A
d
a
p
ti
v
e
-
n
e
u
ro
-
f
u
z
z
y
in
f
e
re
n
c
e
s
y
ste
m
a
p
p
ro
a
c
h
f
o
r
tran
sm
issio
n
li
n
e
f
a
u
lt
c
las
si
f
ica
ti
o
n
a
n
d
l
o
c
a
ti
o
n
in
c
o
r
p
o
ra
ti
n
g
e
ff
e
c
ts
o
f
p
o
w
e
r
s
w
in
g
s
”
,
IET
Ge
n
e
ra
ti
o
n
,
T
ra
n
sm
issio
n
,
D
istrib
u
ti
o
n
,
V
o
l.
2
,
p
p
.
2
3
5
–
2
4
4
,
2
0
0
8
.
[7
1
]
X
iao
m
in
g
Zh
a
o
,
Xu
g
a
n
S
h
i
,
S
h
iq
in
g
Zh
a
n
g
,
“
F
a
c
ial
Ex
p
re
ss
io
n
Re
c
o
g
n
it
io
n
v
ia
De
e
p
L
e
a
rn
in
g
”
,
IET
E
T
e
c
h
n
ica
l
Rev
iew
,
V
o
l
.
3
2
Vo
l.
3
2
,
Iss
.
5,
2
0
1
5
.
[7
2
]
A
.
V
.
Ne
f
ian
a
n
d
M
.
H.
Ha
y
e
s,
“
A
Hid
d
e
n
M
a
rk
o
v
M
o
d
e
l
-
Ba
se
d
A
p
p
ro
a
c
h
f
o
r
F
a
c
e
Re
c
o
g
n
it
io
n
”
,
Pro
c
e
e
d
in
g
s
o
f
IEE
E
In
t.
Co
n
f.
o
n
Aco
u
stic,
S
p
e
e
c
h
a
n
d
S
ig
n
a
l
Pro
c
e
ss
in
g
,
V
o
l
.
5
,
p
p
.
2
7
2
1
-
2
7
2
4
,
1
9
9
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.