I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
201
7
,
p
p
.
3
41
1
~
3
4
2
0
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ec
e.
v
7
i6
.
p
p
3
41
1
-
342
0
3411
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
JE
C
E
Asy
m
m
e
trical
H
a
lf
-
J
o
in
M
ethod
on
D
ua
l
V
isio
n
F
a
ce
R
ecog
nition
E
dy
Wina
rno
1
,
I
m
a
m
H
us
n
i A
l A
m
in
2
,
W
iw
ien H
a
dik
urn
ia
w
a
t
i
3
F
a
c
u
lt
y
o
f
In
f
o
r
m
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
,
Un
iv
e
rsita
s S
ti
k
u
b
a
n
k
,
S
e
m
a
ra
n
g
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
y
9
,
2
0
1
7
R
ev
i
s
ed
J
u
n
2
6
,
2
0
1
7
A
cc
ep
ted
J
u
l
19
,
2
0
1
7
T
h
is
re
se
a
rc
h
p
ro
p
o
se
s
a
m
o
d
e
l
o
f
f
a
c
e
re
c
o
g
n
it
io
n
u
si
n
g
t
h
e
m
e
th
o
d
o
f
jo
in
i
n
g
tw
o
f
a
c
e
i
m
a
g
e
s
f
ro
m
le
f
t
a
n
d
r
ig
h
t
len
s
f
ro
m
a
ste
re
o
v
isio
n
c
a
m
e
ra
n
a
m
e
l
y
h
a
l
f
-
jo
in
m
e
th
o
d
.
Ha
l
f
-
jo
in
m
e
th
o
d
is
u
se
d
i
n
f
a
c
e
ima
g
e
n
o
rm
a
li
z
a
ti
o
n
p
ro
c
e
ss
in
g
.
T
h
e
p
ro
p
o
se
d
h
a
lf
-
jo
in
m
e
th
o
d
is
a
f
a
c
e
i
m
a
g
e
s
jo
in
i
n
g
m
o
d
e
l,
w
h
ich
is
c
a
ll
e
d
a
s
y
m
m
e
tri
c
a
l
h
a
lf
-
jo
in
.
I
n
a
sy
m
m
e
tri
c
a
l
h
a
lf
-
jo
in
m
e
th
o
d
,
a
Ro
I
(re
g
io
n
o
f
i
n
tere
st)
o
f
fa
c
e
i
m
a
g
e
f
ro
m
l
e
f
t
a
n
d
rig
h
t
len
se
s
a
re
p
ro
v
id
e
d
b
a
se
d
o
n
a
x
is
c
e
n
ter
o
f
e
a
c
h
e
y
e
in
e
y
e
d
e
te
c
ti
o
n
.
T
h
e
c
ro
p
p
i
n
g
o
f
f
a
c
e
i
m
a
g
e
f
ro
m
a
s
y
m
m
e
tri
c
a
l
h
a
lf
-
jo
in
m
o
d
e
l
h
a
s
d
if
fe
re
n
t
w
id
th
d
e
p
e
n
d
s
o
n
e
y
e
s
c
o
o
rd
in
a
t
e
lo
c
a
ti
o
n
.
T
h
e
p
r
o
p
o
se
d
sy
ste
m
sh
o
w
s
th
a
t
th
e
a
s
y
m
m
e
tri
c
a
l
h
a
lf
-
jo
in
m
e
th
o
d
c
a
n
p
ro
d
u
c
e
a
b
e
tt
e
r
o
f
f
a
c
e
re
c
o
g
n
it
io
n
ra
te.
T
h
e
e
x
p
e
ri
m
e
n
tal
re
su
lt
s
sh
o
w
th
a
t
th
e
a
s
y
m
m
e
tri
c
a
l
h
a
lf
-
jo
in
m
e
th
o
d
h
a
s
a
b
e
tt
e
r
re
c
o
g
n
it
io
n
ra
te
a
n
d
c
o
m
p
u
tatio
n
ti
m
e
th
a
n
sin
g
le
v
isi
o
n
m
e
th
o
d
a
n
d
sy
m
m
e
tri
c
a
l
h
a
lf
-
jo
in
m
e
th
o
d
.
K
ey
w
o
r
d
:
Face
r
ec
o
g
n
itio
n
As
y
m
m
etr
ical
Half
-
j
o
in
Ster
eo
v
is
io
n
Co
p
y
rig
h
t
©
2
0
1
7
In
stit
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
:
E
d
y
W
i
n
ar
n
o
,
Facu
lt
y
o
f
I
n
f
o
r
m
a
tio
n
T
ec
h
n
o
lo
g
y
,
Un
i
v
er
s
ita
s
Sti
k
u
b
a
n
k
Se
m
ar
a
n
g
,
J
l.
T
r
i L
o
m
b
a
J
u
a
n
g
No
.
1
Sem
ar
an
g
,
5
0
2
4
1
,
J
aw
a
T
en
g
a
h
,
I
n
d
o
n
esia.
E
m
ail:
ed
y
w
in
@
ed
u
.
u
n
is
b
a
n
k
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
I
n
f
ac
e
r
ec
o
g
n
i
tio
n
s
y
s
te
m
,
th
e
u
s
e
o
f
a
ca
m
er
a
as
a
f
a
ce
i
m
a
g
e
ac
q
u
i
s
itio
n
m
et
h
o
d
p
la
y
s
a
n
i
m
p
o
r
tan
t
r
o
le
.
T
h
e
ca
m
er
a
i
s
u
s
ed
to
ca
p
t
u
r
e
t
h
e
i
m
a
g
e
o
f
t
h
e
f
ac
e
u
s
in
g
f
ac
e
d
etec
tio
n
w
h
ic
h
t
h
e
n
i
s
p
r
o
ce
s
s
ed
u
s
in
g
n
o
r
m
al
izatio
n
tech
n
iq
u
es,
f
ea
t
u
r
e
ex
tr
ac
tio
n
an
d
class
if
ica
tio
n
to
id
en
ti
f
y
a
f
ac
e.
I
n
th
e
f
ac
e
r
ec
o
g
n
itio
n
s
y
s
te
m
,
t
h
e
ca
m
e
r
a
is
u
s
ed
to
g
en
er
ate
f
ac
ial
i
m
ag
e
d
ata
to
b
e
p
r
o
ce
s
s
ed
in
th
e
i
m
ag
e
o
f
t
w
o
-
d
i
m
en
s
io
n
al
(
2
D)
o
r
th
r
ee
-
d
im
en
s
io
n
a
l
i
m
a
g
es
(
3
D)
.
Gen
er
all
y
,
to
p
r
o
d
u
ce
a
tw
o
-
d
i
m
e
n
s
io
n
al
i
m
ag
e
o
n
f
ac
e
r
ec
o
g
n
itio
n
w
e
s
till
u
s
e
a
s
i
n
g
le
v
i
s
io
n
ca
m
er
a
f
o
r
t
h
e
a
cq
u
is
itio
n
.
I
n
th
e
3
D
i
m
a
g
e
ac
q
u
is
itio
n
m
o
d
el
s
,
g
en
er
all
y
w
e
u
s
e
s
ter
eo
v
i
s
io
n
ca
m
er
a
o
r
o
th
er
i
m
ag
e
ac
q
u
i
s
itio
n
d
ev
ice
s
s
u
c
h
as
s
ca
n
n
er
e
tc.
Ster
eo
v
is
io
n
ca
m
er
a
ca
n
al
s
o
b
e
u
s
ed
to
g
en
er
ate
b
o
th
2
D
an
d
3
D
i
m
ag
e
s
s
i
m
u
ltan
eo
u
s
l
y
.
A
co
m
b
i
n
atio
n
o
f
2
D
an
d
3
D
f
a
cial
i
m
a
g
e
is
u
s
ed
f
o
r
f
ac
e
r
ec
o
g
n
itio
n
m
o
d
els.
R
esear
c
h
th
a
t
co
m
b
in
es
2
D
a
n
d
3
D
f
ac
ial
i
m
ag
e
s
in
a
f
ac
e
r
e
co
g
n
itio
n
s
y
s
te
m
i
s
ap
p
lied
b
y
[
1
]
an
d
[
2
]
.
I
n
s
o
m
e
s
t
u
d
ies
o
f
f
ac
e
r
ec
o
g
n
i
tio
n
s
y
s
te
m
,
s
ter
eo
v
is
io
n
ca
m
er
a
is
w
id
el
y
u
s
ed
i
n
f
ac
e
r
ec
o
g
n
i
tio
n
p
r
o
ce
s
s
i
n
g
to
p
r
o
d
u
ce
a
3
D
i
m
a
g
e.
T
h
i
s
m
o
d
el
is
u
s
ed
to
r
ec
o
n
s
tr
u
ct
p
o
in
ts
o
f
f
ac
ial
i
m
a
g
e,
a
n
d
t
h
en
u
s
ed
to
g
en
er
ate
a
3
D
f
ac
ial
i
m
ag
e
in
f
ac
e
r
ec
o
g
n
itio
n
.
So
m
e
s
tu
d
ie
s
u
s
i
n
g
s
ter
eo
v
i
s
io
n
ca
m
er
a
is
u
s
e
d
to
g
e
n
er
ate
a
3
D
i
m
a
g
e
r
ec
o
n
s
tr
u
ctio
n
i
n
f
ac
ial
r
ec
o
g
n
itio
n
m
o
d
els,
as c
o
n
d
u
c
ted
b
y
[
3
]
-
[
5
]
.
Fro
m
s
e
v
er
al
p
r
ev
io
u
s
s
t
u
d
ies,
th
e
u
s
e
o
f
s
ter
eo
v
is
io
n
ca
m
er
a
h
as
n
o
t b
ee
n
w
id
el
y
u
s
ed
to
p
r
o
d
u
ce
a
j
o
in
in
g
o
f
2
D
f
ac
ia
l
i
m
ag
e
f
r
o
m
t
h
e
r
ig
h
t
a
n
d
lef
t
ca
m
er
as.
Ster
eo
v
is
io
n
ca
m
er
a
i
s
u
s
ed
t
o
p
r
o
d
u
ce
a
m
o
d
el
o
f
f
ac
e
r
ec
o
g
n
itio
n
in
r
o
b
o
tic
v
is
io
n
s
y
s
te
m
s
.
T
h
e
u
s
e
o
f
s
t
er
eo
v
is
io
n
ca
m
er
a
o
n
a
r
o
b
o
t
ic
v
is
io
n
s
y
s
te
m
i
s
u
s
ed
to
p
er
f
o
r
m
a
f
ac
e
r
ec
o
g
n
itio
n
m
o
d
el
an
d
to
m
ea
s
u
r
e
th
e
d
is
tan
ce
b
et
w
ee
n
th
e
f
a
ce
an
d
th
e
ca
m
er
a.
Ster
eo
v
is
io
n
ca
m
er
a
ca
n
al
s
o
b
e
u
s
ed
to
d
eter
m
in
e
t
h
e
d
is
tan
ce
o
f
a
p
er
s
o
n
's
f
ac
e
u
s
in
g
a
tr
ian
g
u
latio
n
m
et
h
o
d
th
at
u
t
ilizes a
p
air
o
f
c
a
m
er
as to
d
eter
m
i
n
e
th
e
d
is
tan
ce
as c
o
n
d
u
cted
b
y
[
6
]
an
d
[
7
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
2
0
1
7
:
3
4
1
1
–
3
4
2
0
3412
A
lt
h
o
u
g
h
s
o
m
e
s
t
u
d
ies
s
ti
ll
u
s
e
s
ter
eo
v
is
io
n
ca
m
er
a
to
r
ec
o
n
s
tr
u
ct
a
3
D
i
m
a
g
e
f
o
r
f
ac
e
r
e
co
g
n
itio
n
an
d
to
d
eter
m
i
n
e
th
e
d
is
ta
n
ce
o
f
th
e
f
ac
e,
b
u
t
t
h
er
e
is
s
till
a
r
esear
ch
-
g
ap
i
n
r
esear
ch
o
f
2
D
i
m
a
g
e
t
h
at
ca
n
b
e
elab
o
r
ated
u
s
in
g
s
ter
eo
v
i
s
io
n
ca
m
er
a.
I
n
t
h
i
s
r
esear
ch
,
we
cr
ea
te
a
m
o
d
el
to
d
eter
m
i
n
e
th
e
co
r
r
elatio
n
o
f
s
ter
eo
v
is
io
n
ca
m
er
a
th
a
t
elab
o
r
ate
s
th
e
m
o
d
el
o
f
f
ac
e
r
ec
o
g
n
itio
n
u
s
i
n
g
a
2
D
im
a
g
e.
Face
r
ec
o
g
n
itio
n
s
y
s
te
m
u
s
i
n
g
2
D
i
m
a
g
e
r
eq
u
ir
es
les
s
c
o
m
p
u
ti
n
g
t
i
m
e
t
h
a
n
th
e
f
ac
ial
r
ec
o
g
n
itio
n
u
s
in
g
3
D
i
m
a
g
e
[
8
]
-
[1
2
]
.
Face
r
ec
o
g
n
itio
n
u
s
i
n
g
s
ter
eo
v
i
s
io
n
ca
m
er
a
w
ill
p
r
o
d
u
ce
t
w
o
f
ac
e
i
m
ag
e
s
w
i
th
d
if
f
er
e
n
t
a
n
g
les.
B
o
th
o
f
f
ac
e
i
m
a
g
e
h
as
a
d
eta
iled
f
ac
e
i
m
a
g
e
o
n
ea
ch
s
id
e.
L
e
f
t
-
s
id
e
f
ac
e
i
m
a
g
e
o
n
th
e
le
f
t
ca
m
er
a
w
il
l
lo
o
k
m
o
r
e
d
etail
ed
th
a
n
th
e
r
ig
h
t
-
s
i
d
e
f
ac
ial
i
m
a
g
e.
R
ig
h
t
-
s
id
e
f
ac
e
i
m
ag
e
o
n
t
h
e
r
ig
h
t
ca
m
er
a
w
il
l
also
b
e
s
ee
n
i
n
m
o
r
e
d
etail
th
a
n
t
h
e
lef
t
-
s
id
e
f
ac
e
i
m
a
g
e.
T
h
e
im
a
g
e
co
m
b
in
atio
n
b
et
w
ee
n
le
f
t
-
f
ac
e
i
m
ag
e
an
d
r
ig
h
t
-
f
ac
e
i
m
a
g
e
o
n
s
ter
eo
v
i
s
io
n
ca
m
er
a
ca
n
b
e
u
s
ed
to
an
ticip
ate
v
ar
i
atio
n
s
o
f
p
o
s
e
i
n
f
ac
e
r
ec
o
g
n
it
io
n
p
r
o
ce
s
s
[
1
3
]
.
R
eg
ar
d
in
g
t
h
e
i
s
s
u
e
o
f
ill
u
m
i
n
an
ce
,
s
ter
eo
v
i
s
io
n
ca
m
er
a
ca
n
p
r
o
d
u
ce
i
m
a
g
es
with
b
etter
illu
m
i
n
an
ce
le
v
el
s
o
n
ea
ch
o
f
th
e
ca
m
er
a
len
s
.
L
e
f
t
-
s
id
e
f
ac
ial
i
m
ag
e
o
n
t
h
e
le
f
t
ca
m
er
a
w
ill
g
e
t
b
etter
illu
m
i
n
atio
n
t
h
an
t
h
e
r
i
g
h
t
-
s
i
d
e
f
ac
e
i
m
a
g
e.
R
ig
h
t
-
s
id
e
f
ac
e
i
m
a
g
e
o
n
t
h
e
r
i
g
h
t
o
f
t
h
e
c
a
m
er
a
w
il
l
al
s
o
g
e
t
b
etter
illu
m
i
n
atio
n
t
h
an
t
h
e
l
ef
t
-
s
id
e
f
ac
ial
i
m
ag
e.
T
h
e
lef
t
ca
m
er
a
a
n
d
t
h
e
r
i
g
h
t
ca
m
er
a
o
n
s
ter
eo
v
i
s
io
n
ca
m
er
a
ca
n
m
a
x
i
m
ize
t
h
e
le
v
el
o
f
il
lu
m
i
n
a
n
ce
o
n
ea
ch
s
id
e
o
f
th
e
f
ac
e
i
m
a
g
e,
as
w
ell
as
an
ticip
at
in
g
v
ar
iatio
n
o
f
i
llu
m
i
n
a
n
ce
.
T
h
is
s
tu
d
y
p
r
o
p
o
s
es
a
m
o
d
el
o
f
2
D
f
ac
ial
i
m
ag
e
ac
q
u
is
it
io
n
f
r
o
m
t
w
o
ca
m
er
a
len
s
es
o
n
s
ter
eo
v
is
io
n
.
Face
i
m
ag
e
n
o
r
m
a
liza
tio
n
i
s
co
n
d
u
cted
u
s
i
n
g
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
e
s
a
n
d
a
f
a
cial
i
m
ag
e
j
o
in
in
g
m
et
h
o
d
n
a
m
el
y
h
a
lf
-
j
o
in
[
1
3
]
.
Half
-
j
o
in
is
a
j
o
in
in
g
m
o
d
el
b
et
w
ee
n
a
h
al
f
o
f
le
f
t
f
ac
e
i
m
a
g
e
f
r
o
m
lef
t
-
ca
m
er
a
w
it
h
a
h
al
f
o
f
r
ig
h
t
f
ac
e
i
m
a
g
e
f
r
o
m
r
i
g
h
t
-
ca
m
er
a
u
s
i
n
g
s
te
r
eo
v
is
io
n
ca
m
er
a.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
o
f
h
a
lf
-
j
o
in
th
at
u
s
ed
i
n
t
h
is
s
t
u
d
y
is
a
f
ac
e
i
m
a
g
es
j
o
in
in
g
m
o
d
el,
n
a
m
el
y
a
s
y
m
m
etr
ical
h
al
f
-
j
o
in
.
W
e
u
s
e
n
o
r
m
aliza
t
io
n
p
r
o
ce
s
s
o
f
th
e
f
ac
e
i
m
a
g
e
u
s
in
g
p
r
ep
r
o
c
ess
i
n
g
a
n
d
as
y
m
m
etr
ical
h
a
lf
-
j
o
in
m
et
h
o
d
.
I
n
as
y
m
m
etr
ical
h
al
f
-
j
o
in
,
w
e
u
s
e
R
o
I
(
r
eg
io
n
o
f
in
ter
est)
o
f
f
a
ce
i
m
ag
e
f
r
o
m
lef
t
a
n
d
r
ig
h
t
l
en
s
e
s
th
at
p
r
o
v
id
ed
f
r
o
m
a
x
is
ce
n
ter
o
f
ea
c
h
e
y
e
in
e
y
e
d
etec
tio
n
.
I
n
t
h
e
ex
p
er
i
m
en
t
r
esu
l
ts
,
w
e
co
m
p
ar
e
th
e
r
ec
o
g
n
itio
n
r
ate
u
s
i
n
g
th
e
i
m
a
g
es
f
r
o
m
s
i
n
g
le
v
is
io
n
ca
m
er
a,
s
y
m
m
etr
ical
h
a
lf
-
j
o
in
[
1
3
]
an
d
asy
m
m
etr
ical
h
al
f
-
j
o
in
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
As
y
m
m
etr
ical
h
a
lf
-
j
o
in
m
et
h
o
d
is
in
s
p
ir
ed
b
y
th
e
m
o
d
el
o
f
t
h
e
h
u
m
a
n
p
er
s
p
ec
tiv
e
u
s
i
n
g
t
w
o
e
y
e
s
to
s
ee
a
n
d
id
en
ti
f
y
a
p
er
s
o
n
.
B
o
t
h
o
f
t
h
e
h
u
m
a
n
e
y
e
s
ar
e
r
ep
r
es
en
ted
u
s
i
n
g
a
d
u
al
ca
m
er
a
v
i
s
i
o
n
t
h
at
h
as
le
f
t a
n
d
r
ig
h
t
le
n
s
e
s
.
T
h
e
u
s
e
o
f
t
w
o
im
ag
e
s
in
f
ac
e
r
ec
o
g
n
itio
n
is
u
s
ed
to
im
p
le
m
en
t
t
h
e
u
s
e
o
f
t
h
e
f
ea
tu
r
es
o
f
t
h
e
t
w
o
d
if
f
er
e
n
t
v
ie
w
s
s
i
m
u
lta
n
eo
u
s
l
y
.
T
h
is
m
et
h
o
d
co
m
b
i
n
es
t
h
e
lef
t
h
al
f
o
f
t
h
e
i
m
ag
e
a
n
d
th
e
r
ig
h
t
h
alf
o
f
th
e
i
m
a
g
e
in
to
an
i
m
a
g
e
th
at
i
s
r
ea
d
y
to
b
e
ex
tr
ac
ted
o
n
its
f
ea
t
u
r
es.
Face
r
ec
o
g
n
it
io
n
p
r
o
ce
s
s
u
s
i
n
g
as
y
m
m
etr
ical
h
al
f
-
j
o
in
m
eth
o
d
ca
n
b
e
ill
u
s
tr
ated
in
Fig
u
r
e
1
.
Fig
u
r
e
1.
Face
R
ec
o
g
n
itio
n
U
s
in
g
As
y
m
m
etr
ical
Ha
lf
-
j
o
in
M
eth
o
d
I
n
i
m
a
g
e
ac
q
u
i
s
itio
n
p
r
o
ce
s
s
,
f
ac
e
i
m
ag
e
i
s
tak
e
n
b
y
t
h
e
le
f
t
an
d
r
ig
h
t
len
s
es
i
n
s
ter
eo
v
i
s
i
o
n
ca
m
er
a.
T
h
is
p
r
o
ce
s
s
p
r
o
d
u
ce
s
t
w
o
R
GB
i
m
a
g
e
s
ize
o
f
6
4
0
x
4
8
0
p
ix
els.
T
h
e
n
ex
t
s
tep
is
to
d
e
tect
f
ac
e
s
a
n
d
e
y
e
s
u
s
ed
to
a
s
ce
r
tain
w
h
et
h
er
th
er
e
is
a
f
ac
e
i
n
t
h
e
i
m
a
g
e
o
f
t
h
e
ac
q
u
is
itio
n
.
Hal
f
-
j
o
in
s
y
s
te
m
w
il
l
w
o
r
k
i
f
b
o
th
o
f
th
e
le
f
t
a
n
d
r
ig
h
t
le
n
s
e
s
ca
p
tu
r
e
h
u
m
a
n
f
ac
e
i
m
ag
e
s
.
Hal
f
-
j
o
in
p
r
o
ce
s
s
in
g
w
i
ll
n
o
t
w
o
r
k
i
f
o
n
l
y
o
n
e
len
s
t
h
at
ca
p
tu
r
es
th
e
i
m
a
g
e
o
f
a
f
ac
e.
Half
-
j
o
in
p
r
o
ce
s
s
in
g
w
il
l
r
ep
ea
t
u
n
til
t
h
e
ac
q
u
i
s
itio
n
o
f
t
h
e
len
s
e
s
ca
p
tu
r
e
t
h
e
i
m
a
g
e
o
f
a
f
ac
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
A
s
ymm
etri
ca
l H
a
lf
-
Jo
in
Meth
o
d
o
n
Du
a
l V
is
io
n
F
a
ce
R
ec
o
g
n
itio
n
(
E
d
y
W
in
a
r
n
o
)
3413
I
n
as
y
m
m
etr
ical
h
al
f
-
j
o
in
m
et
h
o
d
,
ey
e
d
etec
tio
n
is
u
s
ed
to
lo
ca
te
th
e
ax
is
o
n
th
e
le
f
t
an
d
r
ig
h
t
e
y
e
i
n
th
e
i
m
ag
e.
T
h
e
p
o
s
itio
n
o
f
th
e
e
y
e
is
u
s
ed
as
a
d
eter
m
in
a
n
t
o
f
th
e
lo
ca
tio
n
o
f
t
h
e
cu
t
tin
g
a
r
ea
o
n
th
e
f
ac
e.
T
h
e
p
o
s
itio
n
o
f
th
e
ax
is
is
a
r
elativ
e
p
o
s
itio
n
o
f
t
h
e
r
a
w
i
m
a
g
e
ac
q
u
is
i
tio
n
r
esu
lts
.
T
h
e
n
ex
t
ax
i
s
p
o
s
itio
n
is
r
ec
alcu
lated
in
o
r
d
er
to
illu
s
tr
a
te
th
e
r
elativ
e
p
o
s
it
io
n
o
f
t
h
e
f
ac
e
i
m
ag
e.
As
y
m
m
et
r
ical
h
a
lf
-
j
o
in
m
et
h
o
d
r
eq
u
ir
es
th
e
ca
lcu
latio
n
to
d
eter
m
i
n
e
th
e
p
o
in
t
-
cu
t
tin
g
o
f
t
h
e
i
m
a
g
e.
P
o
in
t
-
cu
tti
n
g
o
f
th
e
i
m
ag
e
is
n
o
t
i
n
t
h
e
m
id
d
le
o
f
t
h
e
i
m
a
g
e.
I
t
w
ill
p
r
o
d
u
ce
th
e
i
m
ag
e
s
w
it
h
d
if
f
er
en
t
w
id
t
h
.
T
h
e
lef
t
s
id
e
ar
ea
o
f
t
h
e
f
ac
e
i
m
ag
e
o
n
th
e
le
f
t
le
n
s
is
t
h
en
c
o
m
b
i
n
ed
w
i
th
th
e
r
ig
h
t
s
id
e
ar
ea
o
f
t
h
e
i
m
a
g
e
o
n
th
e
r
ig
h
t
len
s
.
T
h
e
r
esu
lt
o
f
th
e
f
ac
e
i
m
a
g
e
co
m
b
in
at
io
n
h
a
s
a
d
if
f
er
en
t
w
id
t
h
w
it
h
f
ac
e
i
m
ag
e
s
o
u
r
ce
.
I
t
w
ill
b
e
r
esized
to
th
e
s
a
m
e
w
id
t
h
as th
e
f
ac
e
i
m
a
g
e
s
o
u
r
ce
.
As
y
m
m
etr
ical
h
al
f
-
j
o
in
is
a
m
e
th
o
d
u
s
ed
to
d
iv
id
e
th
e
w
id
t
h
o
f
th
e
f
ac
e
i
m
ag
e
o
n
t
h
e
lef
t
an
d
r
ig
h
t
len
s
ca
m
er
a
w
i
th
u
n
eq
u
al
s
iz
e.
I
n
th
is
m
e
th
o
d
,
Haa
r
ey
e
d
etec
tio
n
[
1
4
]
w
a
s
u
s
ed
to
d
eter
m
in
e
R
o
I
o
f
ey
e
i
m
a
g
e.
T
h
e
f
ir
s
t step
i
n
t
h
e
m
e
th
o
d
o
f
as
y
m
m
etr
ical
h
alf
-
j
o
in
w
a
s
to
d
eter
m
i
n
e
t
h
e
w
id
t
h
o
f
th
e
f
ac
e
i
m
a
g
e
(
w
/
w
id
t
h
)
in
p
ix
e
ls
.
T
h
is
m
et
h
o
d
r
eq
u
ir
es
th
e
m
id
p
o
in
t
o
f
t
h
e
lef
t
e
y
e
(
le
)
an
d
r
ig
h
t
e
y
e
(
re
)
to
d
eter
m
in
e
t
h
e
p
o
in
t
o
f
cu
ttin
g
(
c
)
.
T
h
e
p
o
in
t
o
f
c
u
tt
in
g
i
s
th
e
s
u
m
o
f
t
h
e
m
id
p
o
in
t
o
f
th
e
le
f
t
a
n
d
r
i
g
h
t
e
y
es.
T
h
e
n
,
d
iv
id
ed
b
y
2
(
c
= (
le
+
re
)
/ 2
)
.
I
m
ag
e
c
u
tti
n
g
p
o
in
t
is
ca
lc
u
lat
ed
o
n
th
e
le
f
t
a
n
d
r
ig
h
t
i
m
a
g
es
,
s
o
th
at
t
h
e
c
u
tti
n
g
p
o
in
t
o
f
t
h
e
lef
t
a
n
d
r
ig
h
t
i
m
a
g
es
w
i
ll
b
e
d
if
f
er
en
t.
Face
i
m
a
g
e
o
f
t
h
e
lef
t
le
n
s
o
f
th
e
ca
m
er
a
is
t
h
e
cr
o
p
p
in
g
a
r
ea
b
et
w
ee
n
t
h
e
le
f
t
p
o
in
t
(
x
=
0
)
an
d
th
e
p
ix
el
s
b
ef
o
r
e
th
e
cu
t
tin
g
p
o
in
t
(
x
=
c
-
1
)
.
Face
i
m
a
g
e
o
f
th
e
r
i
g
h
t
len
s
ca
m
er
a
is
th
e
cr
o
p
p
in
g
ar
ea
b
etw
ee
n
th
e
c
u
t
tin
g
p
o
in
t
(
x
=
c
)
an
d
th
e
r
ig
h
t
-
m
o
s
t
p
o
in
t
o
f
th
e
i
m
ag
e
(
x
=
w
-
1
)
.
B
o
th
o
f
f
ac
e
i
m
a
g
es
ar
e
th
e
n
co
m
b
i
n
ed
an
d
r
esized
s
i
m
ilar
to
th
e
s
ize
o
f
t
h
e
f
ac
ia
l i
m
a
g
e
(
w
).
T
h
e
f
ir
s
t
s
tep
in
th
e
m
e
th
o
d
o
f
as
y
m
m
etr
ical
h
al
f
-
j
o
in
is
to
d
eter
m
i
n
e
t
h
e
ax
i
s
p
o
in
t
o
f
c
u
tti
n
g
p
o
in
t (
c
l
a
)
.
T
h
is
cu
tti
n
g
p
o
in
t
is
t
h
e
m
i
d
p
o
in
t
b
et
w
ee
n
t
h
e
p
o
in
t
o
f
t
h
e
lef
t
e
y
e
(
le
)
an
d
r
ig
h
t
e
y
e
p
o
i
n
t
(
re
)
o
n
th
e
f
ac
e
i
m
a
g
e
d
etec
tio
n
r
es
u
lt
f
r
o
m
le
f
t le
n
s
o
f
t
h
e
ca
m
er
a,
a
s
s
h
o
w
n
i
n
E
q
u
atio
n
(
1
)
.
c
l
a
=
2
re
le
(
1
)
T
h
e
s
ec
o
n
d
p
r
o
ce
s
s
is
to
d
eter
m
in
e
t
h
e
a
x
is
o
f
cu
t
tin
g
p
o
in
t
(
c
r
a
).
c
r
a
is
t
h
e
m
id
p
o
in
t
b
et
w
ee
n
th
e
p
o
in
t
o
f
th
e
le
f
t
e
y
e
(
le
)
an
d
r
ig
h
t
e
y
e
p
o
in
t
(
re
)
o
n
t
h
e
f
ac
e
i
m
a
g
e
d
etec
tio
n
r
es
u
lts
o
f
th
e
r
ig
h
t
le
n
s
o
f
th
e
ca
m
er
a,
as
s
h
o
w
n
in
E
q
u
a
t
io
n
(
2
)
.
c
r
a
=
2
re
le
(
2
)
T
h
e
n
ex
t
s
tep
is
to
cu
t
t
h
e
f
ac
e
i
m
a
g
e
d
etec
tio
n
r
es
u
lt
o
f
t
h
e
lef
t
le
n
s
o
f
th
e
ca
m
er
a
a
h
j
l
e
f
t
.
C
u
tti
n
g
p
r
o
ce
s
s
is
d
o
n
e
in
t
h
e
ar
ea
o
f
th
e
f
ac
e
(
x
=
0
)
to
(
x
=
c
l
a
-
1
)
as s
h
o
w
n
i
n
E
q
u
atio
n
(
3
)
.
a
hj
l
e
f
t
=
1
0
c
l
a
x
cl
a
(
3
)
T
h
e
s
am
e
p
r
o
ce
s
s
is
al
s
o
p
er
f
o
r
m
ed
o
n
th
e
f
ac
e
i
m
a
g
e
d
etec
tio
n
r
esu
l
ts
o
f
t
h
e
r
ig
h
t
le
n
s
o
f
t
h
e
ca
m
er
a
a
h
j
r
i
g
h
t
.
C
u
tti
n
g
p
r
o
ce
s
s
i
s
d
o
n
e
in
t
h
e
ar
ea
o
f
t
h
e
f
ac
e
(
x
=
c
r
a
)
to
(
x
=
1
w
)
as
s
h
o
w
n
i
n
E
q
u
atio
n
(
4
)
.
a
hj
r
i
gh
t
=
1
w
c
r
a
x
w
(
4
)
T
h
e
r
esu
lt
o
f
a
j
o
in
b
et
w
ee
n
t
h
e
f
ac
e
i
m
a
g
e
cr
o
p
p
ed
o
n
th
e
lef
t
le
n
s
o
f
t
h
e
ca
m
er
a
a
h
j
l
e
f
t
an
d
th
e
f
ac
e
i
m
ag
e
cr
o
p
p
ed
o
n
th
e
r
ig
h
t le
n
s
o
f
t
h
e
ca
m
er
a
a
h
j
r
i
g
h
t
is
s
h
o
wn
in
E
q
u
a
tio
n
(
5
)
an
d
(
6
)
.
a
h
j
=
a
hj
l
e
f
t
+
a
hj
r
i
gh
t
(
5
)
a
h
j
=
1
0
c
l
a
x
cl
a
+
1
w
c
r
a
x
w
(
6
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
2
0
1
7
:
3
4
1
1
–
3
4
2
0
3414
Face
i
m
a
g
e
co
m
b
in
i
n
g
r
esu
l
t
h
as
d
if
f
er
en
t
p
ix
e
l
s
ize
th
a
n
th
e
o
r
ig
in
al
s
ize.
R
esizi
n
g
p
r
o
ce
s
s
is
n
ee
d
ed
to
r
esto
r
e
it.
A
s
y
m
m
et
r
ical
h
alf
-
j
o
in
m
et
h
o
d
h
as
t
wo
ad
d
itio
n
al
R
o
I
d
ete
r
m
i
n
i
n
g
s
tep
to
g
et
a
p
ar
tial
ar
ea
o
f
th
e
f
ac
e
o
n
th
e
le
f
t
an
d
r
ig
h
t
len
s
.
R
o
I
i
s
a
v
er
tical
r
ec
tan
g
u
lar
ar
ea
.
E
ac
h
R
o
I
h
as
d
if
f
er
en
t
w
id
th
b
u
t
h
as
th
e
s
a
m
e
h
ei
g
h
t.
T
h
is
p
r
o
ce
s
s
al
s
o
s
ee
k
s
t
h
e
co
o
r
d
in
ate
o
f
th
e
to
p
le
f
t,
to
p
r
i
g
h
t,
b
o
tto
m
le
f
t
an
d
b
o
tto
m
r
ig
h
t.
T
h
e
co
o
r
d
in
ates,
w
id
t
h
an
d
h
ei
g
h
t
o
f
t
h
e
i
m
a
g
e
u
s
ed
in
t
h
e
ca
lc
u
latio
n
ar
e
t
h
e
co
o
r
d
in
ates
o
f
th
e
R
o
I
d
u
r
in
g
f
ac
e
d
etec
tio
n
.
T
h
e
m
id
p
o
in
t
o
f
th
e
f
ac
e
o
f
t
h
e
a
s
y
m
m
etr
ical
h
al
f
-
j
o
in
i
s
t
h
e
m
id
p
o
in
t
b
et
w
ee
n
t
h
e
le
f
t
an
d
r
ig
h
t
e
y
es.
E
y
e
d
etec
tio
n
u
s
i
n
g
Haa
r
[
1
4
]
w
as
u
s
ed
to
d
etec
t
t
h
e
l
o
ca
tio
n
o
f
t
h
e
le
f
t
an
d
r
i
g
h
t
e
y
es.
E
y
e
d
etec
ti
o
n
is
u
s
ed
to
o
b
tain
th
e
co
o
r
d
in
at
es
o
f
t
h
e
p
o
in
t
)
,
(
a
a
y
x
o
n
th
e
to
p
lef
t
co
r
n
er
o
f
t
h
e
le
f
t
e
y
e
i
m
ag
e,
t
h
e
co
o
r
d
in
ates o
f
p
o
in
ts
)
,
(
a
b
y
x
o
n
th
e
to
p
lef
t c
o
r
n
er
o
f
th
e
r
i
g
h
t e
y
e
i
m
ag
e,
t
h
e
w
id
t
h
(
w
)
an
d
h
ei
g
h
t (
h
).
T
h
e
m
id
p
o
in
t
o
f
t
h
e
f
ac
e
(
c
l
a
an
d
c
r
a
)
is
ca
lcu
lated
b
y
d
i
v
id
i
n
g
t
h
e
s
u
m
o
f
th
e
m
id
p
o
in
t
o
f
t
h
e
e
y
e.
T
h
e
m
id
p
o
in
t
o
f
t
h
e
lef
t
e
y
e
is
th
e
p
o
in
t
o
f
t
h
e
u
p
p
er
lef
t
co
r
n
er
o
f
lef
t
e
y
e
)
,
(
a
a
y
x
p
lu
s
h
al
f
th
e
w
id
t
h
o
f
th
e
le
f
t
e
y
e
i
m
a
g
e
(
½
1
w
)
.
T
h
e
m
id
p
o
in
t
o
f
th
e
r
i
g
h
t
e
y
e
i
s
th
e
p
o
in
t
o
f
th
e
u
p
p
er
lef
t
co
r
n
er
o
f
r
ig
h
t
e
y
e
)
,
(
a
b
y
x
p
lu
s
h
al
f
th
e
w
id
t
h
o
f
th
e
r
ig
h
t
e
y
e
i
m
a
g
e
(
½
2
w
)
.
Dete
r
m
i
n
at
io
n
o
f
co
o
r
d
in
ates
o
f
lef
t
an
d
r
ig
h
t
e
y
es o
n
t
h
e
as
y
m
m
e
tr
ical
h
al
f
-
j
o
in
ca
n
b
e
illu
s
tr
ated
in
Fig
u
r
e
2
.
Fig
u
r
e
2
.
T
h
e
C
o
o
r
d
in
ates o
f
L
e
f
t a
n
d
R
ig
h
t E
y
e
s
o
n
As
y
m
m
etr
ical
H
al
f
-
j
o
i
n
C
o
o
r
d
in
ates
o
f
th
e
s
tar
ti
n
g
p
o
i
n
t
)
,
(
a
a
y
x
,
)
,
(
a
b
y
x
,
1
w
an
d
2
w
is
u
s
ed
to
o
b
tain
t
h
e
v
al
u
e
o
f
t
h
e
m
id
p
o
in
t
o
f
t
h
e
lef
t
e
y
e
(
le
)
,
th
e
m
id
p
o
in
t
o
f
t
h
e
r
i
g
h
t
e
y
e
(
re
)
an
d
t
h
e
m
id
d
le
o
f
cu
tti
n
g
p
o
in
t
i
n
th
e
f
ac
e
i
m
a
g
e
(
c
l
a
an
d
c
r
a
)
.
le
is
ca
lcu
lated
u
s
i
n
g
E
q
u
a
tio
n
(
7
)
.
le
=
a
x
+½
1
w
(
7
)
T
o
d
eter
m
i
n
e
th
e
v
al
u
e
o
f
re
is
s
o
lv
ed
u
s
i
n
g
E
q
u
atio
n
(
8
)
.
re
=
b
x
+½
2
w
(
8
)
T
h
e
m
id
p
o
in
t o
f
th
e
f
ac
e
(
c
l
a
d
an
c
r
a
)
is
ca
lcu
lated
u
s
i
n
g
E
q
u
at
io
n
(
9
)
an
d
(
1
0
)
.
c
l
a
=
½
(
le
+
re
)
o
r
c
r
a
=
½
(
le
+
re
)
(
9
)
c
l
a
=
½
(
a
x
+½
1
w
+
b
x
+½
2
w
)
o
r
c
r
a
=
½
(
a
x
+½
1
w
+
b
x
+½
2
w
)
(
1
0
)
F
a
c
e
I
mag
e
le
L
e
f
t
e
y
e
R
i
g
h
t
e
y
e
re
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
A
s
ymm
etri
ca
l H
a
lf
-
Jo
in
Meth
o
d
o
n
Du
a
l V
is
io
n
F
a
ce
R
ec
o
g
n
itio
n
(
E
d
y
W
in
a
r
n
o
)
3415
c
l
a
an
d
c
r
a
ar
e
u
s
ed
as
t
h
e
c
u
tti
n
g
p
o
in
t
to
d
i
v
id
e
R
o
I
o
f
f
ac
e
i
m
a
g
es
o
n
le
f
t
a
n
d
r
ig
h
t
ar
ea
.
B
o
th
o
f
R
o
I
f
ac
e
i
m
ag
e
s
ar
e
th
e
n
co
m
b
in
e
d
to
p
r
o
d
u
ce
a
f
ac
e
i
m
ag
e
t
h
at
w
ill
b
e
u
s
ed
in
t
h
e
f
ea
tu
r
e
e
x
t
r
ac
tio
n
p
r
o
ce
s
s
.
2
.
1
.
Ro
I
o
f
lef
t
ha
lf
-
f
a
ce
)
,
(
0
0
y
x
an
d
)
,
(
1
0
y
x
ar
e
th
e
co
o
r
d
in
ates
o
f
t
h
e
i
m
a
g
e
p
ix
el
co
l
u
m
n
.
)
,
(
0
1
y
x
an
d
)
,
(
1
1
y
x
ar
e
d
eter
m
in
ed
b
ased
o
n
th
e
cu
tti
n
g
p
o
in
t
(
c
l
a
)
in
t
h
e
i
m
a
g
e,
s
o
1
x
ca
n
b
e
d
i
s
co
v
er
ed
.
1
y
is
d
eter
m
in
ed
b
ased
o
n
th
e
h
e
ig
h
t (
h
)
in
th
e
i
m
a
g
e.
R
o
I
p
ar
tial im
a
g
e
o
f
t
h
e
le
f
t f
ac
e
o
f
th
e
le
f
t le
n
s
ca
n
b
e
ill
u
s
tr
ated
in
Fig
u
r
e
3
.
Fig
u
r
e
3
.
R
o
I
P
ar
tial I
m
ag
e
o
f
T
h
e
L
ef
t Fac
e
o
f
T
h
e
L
ef
t
L
e
n
s
1
x
an
d
1
y
is
ca
lcu
lated
u
s
i
n
g
E
q
u
at
io
n
(
1
1
)
an
d
(
1
2
)
.
1
x
=
0
x
+(
c
l
a
-
1
)
(
1
1
)
1
y
=
0
y
+(
h
-
1
)
(
1
2
)
A
,
E
,
C
an
d
F
ar
e
t
h
e
R
o
I
co
o
r
d
in
ate
p
o
in
ts
o
f
th
e
p
ar
tial
i
m
a
g
e
o
f
t
h
e
le
f
t
f
ac
e
o
n
th
e
lef
t
le
n
s
.
I
t
ca
n
b
e
ca
lcu
lated
u
s
i
n
g
E
q
u
at
io
n
(
1
3
)
-
(
1
6
)
.
A=
)
,
(
0
0
y
x
(
1
3
)
E
=(
0
x
+(
c
l
a
-
1
)
,
0
y
)
(
1
4
)
C
=(
0
x
,
0
y
+(
h
-
1
)
)
(
1
5
)
F=(
0
x
+(
c
l
a
-
1
)
,
0
y
+(
h
-
1
)
)
(
16)
R
o
I
at
co
o
r
d
in
ates
A
,
E
,
C
a
n
d
F
is
t
h
en
co
m
b
i
n
ed
w
it
h
a
p
ar
tial
i
m
ag
e
o
f
th
e
r
ig
h
t
f
ac
e
o
f
t
h
e
r
i
g
h
t
len
s
o
f
th
e
ca
m
er
a.
T
h
is
i
m
a
g
e
is
a
f
ac
e
i
m
ag
e
o
f
as
y
m
m
etr
ic
al
h
al
f
-
j
o
in
.
2
.
2
.
Ro
I
o
f
rig
ht
ha
lf
-
f
a
ce
)
,
(
0
0
y
x
an
d
)
,
(
1
0
y
x
ar
e
th
e
p
r
im
ar
y
co
o
r
d
in
ate
p
o
in
ts
in
th
e
co
lu
m
n
p
i
x
el
o
f
i
m
a
g
e.
T
h
ese
co
o
r
d
in
ate
p
o
in
ts
ar
e
u
s
ed
to
d
eter
m
in
e
th
e
R
OI
o
f
t
h
e
r
i
g
h
t
f
ac
e
i
m
ag
e
o
f
t
h
e
r
ig
h
t
l
en
s
.
)
,
(
0
1
y
x
an
d
)
,
(
1
1
y
x
ar
e
d
eter
m
i
n
ed
b
y
t
h
e
c
u
tti
n
g
p
o
in
t
(
c
r
a
)
.
)
,
(
0
2
y
x
an
d
)
,
(
1
2
y
x
ar
e
d
eter
m
in
ed
b
y
t
h
e
w
id
t
h
(
w
)
B
a
c
k
g
r
o
u
n
d
I
mag
e
C
i
t
r
a
W
a
j
a
h
A
E
C
)
F
h
F
a
c
e
i
mag
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
2
0
1
7
:
3
4
1
1
–
3
4
2
0
3416
an
d
h
ei
g
h
t
(
h
)
o
f
t
h
e
f
ac
e
i
m
ag
e,
an
d
i
t
is
ca
lc
u
lated
f
r
o
m
th
e
co
o
r
d
in
ates
o
f
)
,
(
0
0
y
x
an
d
)
,
(
1
0
y
x
.
R
o
I
p
ar
tial i
m
ag
e
o
f
th
e
r
i
g
h
t
f
ac
e
o
f
th
e
r
ig
h
t le
n
s
ca
n
b
e
illu
s
tr
a
ted
in
Fig
u
r
e
4
.
Fig
u
r
e
4
.
R
o
I
P
ar
tial I
m
ag
e
o
f
T
h
e
R
ig
h
t
Face
o
f
T
h
e
R
i
g
h
t
L
e
n
s
1
x
,
1
y
an
d
2
x
is
ca
lcu
lated
u
s
in
g
E
q
u
atio
n
(
1
7
)
,
(
1
8
)
an
d
(
1
9
)
.
1
x
=
0
x
+
c
r
a
(
1
7
)
1
y
=
0
y
+(
h
-
1
)
(
1
8
)
2
x
=
0
x
+(
w
-
1
)
(
1
9
)
G,
B
,
H
an
d
D
ar
e
t
h
e
R
o
I
co
o
r
d
in
ate
p
o
in
ts
o
f
t
h
e
p
ar
tia
l
i
m
a
g
e
o
f
t
h
e
r
i
g
h
t
f
ac
e
o
n
t
h
e
r
i
g
h
t
le
n
s
.
I
t
is
ca
lcu
lated
u
s
i
n
g
E
q
u
at
io
n
(
2
0
)
-
(
2
3
)
.
G=
(
0
x
+
c
r
a
,
0
y
)
(
2
0
)
B
=(
0
x
+(
w
-
1
)
,
0
y
)
(
2
1
)
H=
(
0
x
+
c
r
a
,
0
y
+(
h
-
1
)
)
(
2
2
)
D=
(
0
x
+(
w
-
1
)
,
0
y
+(
h
-
1
)
)
(
2
3
)
R
o
I
at
co
o
r
d
in
ates
G,
B
,
H
an
d
D
is
th
e
n
co
m
b
i
n
ed
w
it
h
R
o
I
at
co
o
r
d
in
ates
A
,
E
,
C
a
n
d
F.
I
t
is
r
e
s
ized
to
1
6
0
x
1
6
0
p
ix
el
o
f
i
m
a
g
e.
T
h
is
i
m
a
g
e
is
a
f
ac
e
i
m
a
g
e
o
f
as
y
m
m
etr
ical
h
alf
-
j
o
in
.
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
T
h
is
s
t
u
d
y
u
s
es
t
h
e
Vio
la
-
J
o
n
es
f
ac
e
d
etec
tio
n
.
E
ac
h
le
f
t a
n
d
r
ig
h
t le
n
s
e
s
o
f
a
s
ter
eo
v
is
io
n
ca
m
er
a
i
s
u
s
ed
to
tak
e
a
r
a
w
i
m
a
g
e
w
i
t
h
t
h
e
b
ac
k
g
r
o
u
n
d
i
m
a
g
e
s
ize
o
f
6
4
0
x
4
8
0
p
ix
els
an
d
th
e
d
e
tecte
d
f
ac
e
i
m
ag
e.
W
id
th
an
d
h
eig
h
t
o
f
t
h
e
f
ac
e
i
m
a
g
e
d
ep
en
d
s
o
n
th
e
d
is
ta
n
ce
b
et
w
ee
n
th
e
f
ac
e
a
n
d
t
h
e
ca
m
er
a.
T
h
e
f
ac
e
i
m
a
g
e
d
etec
ted
o
n
th
e
b
ac
k
g
r
o
u
n
d
i
m
a
g
e
is
m
ar
k
ed
w
it
h
t
h
e
co
o
r
d
in
ates
o
f
th
e
p
o
s
itio
n
o
f
th
e
f
ac
e
i
m
a
g
e
R
o
I
.
T
h
e
f
ac
e
i
m
ag
e
R
o
I
is
th
e
n
s
to
r
ed
i
n
to
th
e
v
ar
iab
l
e
s
to
r
ag
e
f
ac
e
i
m
a
g
e.
T
h
e
f
ac
e
i
m
a
g
e
i
s
t
h
e
n
n
o
r
m
alize
d
u
s
i
n
g
p
r
ep
r
o
ce
s
s
in
g
an
d
h
al
f
-
j
o
in
m
et
h
o
d
.
I
n
p
r
ep
r
o
ce
s
s
in
g
,
t
h
e
c
u
tt
in
g
p
r
o
ce
s
s
is
co
n
d
u
cted
to
ta
k
e
th
e
f
ac
e
i
m
a
g
e
an
d
it
i
s
co
n
v
er
ted
in
t
o
g
r
a
y
s
ca
le.
T
h
e
g
r
a
y
s
ca
le
i
m
a
g
e
is
t
h
en
ad
j
u
s
ted
u
s
i
n
g
th
e
r
esizin
g
p
r
o
ce
s
s
to
g
et
a
f
ac
e
i
m
ag
e
s
ize
o
f
B
a
c
k
g
r
o
u
n
d
I
mag
e
F
a
c
e
i
mag
e
H
h
G
A
B
C
)
D
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
A
s
ymm
etri
ca
l H
a
lf
-
Jo
in
Meth
o
d
o
n
Du
a
l V
is
io
n
F
a
ce
R
ec
o
g
n
itio
n
(
E
d
y
W
in
a
r
n
o
)
3417
1
6
0
x
1
6
0
p
ix
els.
T
h
e
ad
j
u
s
t
m
e
n
t
o
f
co
n
tr
ast
an
d
b
r
ig
h
t
n
es
s
i
s
co
n
d
u
cted
u
s
i
n
g
t
h
e
h
is
to
g
r
a
m
eq
u
al
izatio
n
.
I
t
is
u
s
ed
to
p
r
o
d
u
ce
th
e
b
etter
f
ac
e
i
m
a
g
e
q
u
ali
t
y
.
I
m
a
g
e
r
es
u
lts
in
ea
ch
o
f
t
h
e
ca
m
er
a
le
n
s
is
th
e
n
p
r
o
ce
s
s
ed
u
s
i
n
g
th
e
m
e
th
o
d
o
f
as
y
m
m
etr
ical
h
alf
-
j
o
in
.
T
h
e
ex
a
m
p
les
o
f
s
o
m
e
R
o
I
f
ac
e
i
m
ag
e
p
r
ep
r
o
ce
s
s
i
n
g
r
es
u
lt
s
o
n
a
s
ter
eo
v
is
io
n
ca
m
er
a
i
n
s
o
m
e
p
o
s
es a
n
d
tak
e
n
at
th
e
s
a
m
e
t
i
m
e
as
s
ee
n
i
n
Fi
g
u
r
e
5
.
Fig
u
r
e
5
.
Face
I
m
a
g
e
P
r
ep
r
o
c
ess
i
n
g
R
es
u
lt
s
o
n
A
Ster
eo
Vi
s
io
n
C
a
m
er
a,
(
ab
o
v
e)
T
h
e
im
a
g
e
o
f
T
h
e
L
e
f
t
Le
n
s
,
(
b
o
tto
m
)
T
h
e
I
m
a
g
e
o
f
T
h
e
R
ig
h
t L
e
n
s
L
e
f
t
e
y
e
d
etec
tio
n
an
d
r
i
g
h
t
e
y
e
d
etec
tio
n
is
u
s
ed
in
as
y
m
m
etr
ical
h
al
f
-
j
o
in
m
e
th
o
d
to
d
et
er
m
in
e
th
e
cr
o
p
p
e
d
ar
ea
o
f
f
ac
e
i
m
a
g
e.
T
h
e
lef
t
f
ac
e
i
m
a
g
e
is
p
r
o
d
u
ce
d
b
y
co
o
r
d
in
ates
A
)
,
(
0
0
y
x
,
E
)
,
(
0
1
y
x
,
C
)
,
(
1
0
y
x
an
d
F
)
,
(
1
1
y
x
.
T
h
e
r
ig
h
t
o
n
e
is
p
r
o
d
u
ce
d
b
y
G
)
,
(
0
1
y
x
,
B
)
,
(
0
2
y
x
,
H
)
,
(
1
1
y
x
d
an
D
)
,
(
1
2
y
x
.
T
h
e
s
ize
o
f
all
o
f
th
e
f
ac
e
i
m
a
g
es
h
as d
if
f
er
e
n
t
w
id
th
a
n
d
h
e
ig
h
t.
T
h
e
p
ar
tial
i
m
ag
e
o
f
t
h
e
le
f
t
f
ac
e
o
f
th
e
lef
t
le
n
s
a
n
d
th
e
p
a
r
tial
i
m
a
g
e
o
f
th
e
r
ig
h
t
f
ac
e
o
f
t
h
e
r
ig
h
t
len
s
ar
e
t
h
en
co
m
b
i
n
ed
to
p
r
o
d
u
ce
a
f
ac
e
i
m
ag
e
o
f
as
y
m
m
e
tr
ical
h
al
f
-
j
o
in
.
T
h
e
ex
a
m
p
les
o
f
f
ac
e
i
m
ag
e
s
o
f
as
y
m
m
etr
ical
h
a
lf
-
j
o
in
w
i
t
h
d
if
f
er
en
t size
i
s
s
ee
n
in
Fig
u
r
e
6
.
Fig
u
r
e
6
.
T
h
e
Face
I
m
a
g
e
o
f
As
y
m
m
etr
ical
Hal
f
-
J
o
in
,
(
ab
o
v
e)
T
h
e
P
ar
tial I
m
ag
e
o
f
T
h
e
L
e
f
t Fac
e
o
f
T
h
e
L
e
f
t
L
en
s
,
(
m
id
le)
T
h
e
P
ar
tial
I
m
ag
e
o
f
T
h
e
R
i
g
h
t Fac
e
o
f
T
h
e
R
i
g
h
t L
e
n
s
,
(
b
o
tto
m
)
C
o
m
b
in
ed
Mo
d
el
I
n
th
i
s
r
esear
ch
,
th
e
te
s
t
is
co
n
d
u
cted
b
y
co
m
p
ar
in
g
b
et
w
ee
n
d
ataset
f
r
o
m
s
in
g
le
v
is
io
n
f
ac
e
i
m
ag
e
,
s
y
m
m
etr
ical
h
a
lf
-
j
o
in
f
ac
e
i
m
ag
e
(
d
iv
id
in
g
s
ize
o
f
f
ac
e
i
m
ag
e
s
y
m
m
etr
ical
l
y
)
[
1
3
]
an
d
as
y
m
m
etr
ical
h
a
lf
-
j
o
in
f
ac
e
i
m
a
g
e.
T
h
e
ex
a
m
p
les
o
f
all
o
f
th
e
i
m
a
g
es
i
s
s
ee
n
in
Fig
u
r
e
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
2
0
1
7
:
3
4
1
1
–
3
4
2
0
3418
Fig
u
r
e
7
.
Face
I
m
a
g
e
s
o
f
T
esti
n
g
,
(
le
f
t)
Si
n
g
le
V
is
io
n
Face
I
m
ag
e,
(
m
id
le)
Face
I
m
a
g
e
o
f
S
y
m
m
etr
ical
Hal
f
-
J
o
in
,
(
r
ig
h
t)
Face
I
m
a
g
e
o
f
A
s
y
m
m
etr
ica
l H
alf
-
J
o
in
T
est
f
ac
e
i
m
a
g
es
ar
e
tak
e
n
f
r
o
m
4
0
p
eo
p
le
ea
ch
w
er
e
ta
k
en
1
0
ti
m
e
s
.
T
h
e
r
esu
l
ts
ar
e
4
0
0
s
in
g
le
v
is
io
n
f
ac
e
i
m
a
g
es,
4
0
0
s
y
m
m
etr
ical
h
alf
-
j
o
in
f
ac
e
i
m
a
g
es
an
d
4
0
0
asy
m
m
etr
ical
h
al
f
-
j
o
in
f
ac
e
i
m
a
g
e.
T
h
e
test
i
n
g
i
s
co
n
d
u
c
ted
b
y
co
m
p
ar
in
g
ea
ch
test
i
m
a
g
es
w
i
th
th
e
tr
ain
in
g
i
m
ag
e
s
at
ea
c
h
m
et
h
o
d
.
I
t
is
also
co
n
d
u
cted
b
y
co
m
p
ar
i
n
g
s
i
n
g
l
e
v
is
io
n
te
s
t
i
m
a
g
e
s
w
ith
s
in
g
l
e
v
is
io
n
tr
ai
n
i
n
g
i
m
a
g
es,
s
y
m
m
etr
ic
h
al
f
-
j
o
in
test
i
m
a
g
es
w
i
th
s
y
m
m
e
tr
ic
h
al
f
-
j
o
in
tr
ain
in
g
i
m
a
g
es
a
n
d
as
y
m
m
etr
ical
h
al
f
-
j
o
in
test
i
m
a
g
es
w
ith
as
y
m
m
etr
ica
l
h
al
f
-
j
o
in
tr
ain
i
n
g
i
m
a
g
es
o
n
r
ea
l
ti
m
e
co
n
d
it
io
n
s
.
T
h
e
tes
t
r
esu
lts
s
h
o
w
t
h
at
th
e
m
eth
o
d
o
f
s
i
n
g
le
v
is
io
n
test
i
n
g
h
a
s
an
av
er
a
g
e
ac
cu
r
ac
y
o
f
9
6
%,
o
n
a
s
y
m
m
e
tr
ical
h
alf
-
j
o
in
m
et
h
o
d
h
as
an
av
er
ag
e
ac
cu
r
ac
y
o
f
9
3
.
2
5
%
an
d
asy
m
m
etr
ical
h
alf
-
j
o
in
m
eth
o
d
h
as
a
n
av
er
ag
e
ac
cu
r
ac
y
o
f
9
7
.
5
%.
T
h
e
av
er
ag
e
o
f
f
ac
e
r
ec
o
g
n
itio
n
r
ate
u
s
i
n
g
a
s
i
n
g
le
v
is
io
n
,
s
y
m
m
etr
ical
a
n
d
as
y
m
m
etr
ical
h
alf
-
j
o
in
is
il
lu
s
tr
ated
in
Fi
g
u
r
e
8
.
Fig
u
r
e
8
.
Face
R
ec
o
g
n
itio
n
R
a
te
(
%)
T
h
e
m
ea
s
u
r
e
m
e
n
t
o
f
co
m
p
u
t
atio
n
ti
m
e
o
f
f
ac
e
r
ec
o
g
n
itio
n
is
co
n
d
u
cted
s
i
m
u
l
tan
eo
u
s
l
y
w
it
h
th
e
test
i
n
g
o
f
f
ac
e
r
ec
o
g
n
it
io
n
r
at
e
o
n
t
h
e
m
et
h
o
d
o
f
s
i
n
g
le
v
is
i
o
n
,
s
y
m
m
etr
ical
an
d
as
y
m
m
et
r
ical
h
al
f
-
j
o
in
.
T
h
e
test
s
p
r
o
d
u
ce
d
an
av
er
a
g
e
o
f
c
o
m
p
u
tatio
n
ti
m
e
o
f
f
ac
e
r
ec
o
g
n
itio
n
o
n
ea
ch
m
et
h
o
d
.
T
h
e
test
r
esu
lts
s
h
o
w
t
h
at
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
A
s
ymm
etri
ca
l H
a
lf
-
Jo
in
Meth
o
d
o
n
Du
a
l V
is
io
n
F
a
ce
R
ec
o
g
n
itio
n
(
E
d
y
W
in
a
r
n
o
)
3419
th
e
s
i
n
g
le
v
i
s
io
n
m
et
h
o
d
h
as
an
av
er
ag
e
co
m
p
u
tat
io
n
ti
m
e
o
f
4
5
.
4
9
m
s
,
s
y
m
m
e
tr
ical
h
al
f
-
j
o
in
m
et
h
o
d
h
as
a
n
av
er
ag
e
co
m
p
u
tatio
n
t
i
m
e
o
f
4
7
.
8
0
m
s
an
d
as
y
m
m
etr
ical
h
alf
-
j
o
in
h
as
a
n
av
er
a
g
e
co
m
p
u
tatio
n
ti
m
e
o
f
4
5
.
9
9
m
s
.
T
h
e
co
m
p
u
tat
io
n
ti
m
e
o
f
f
ac
e
r
ec
o
g
n
itio
n
o
n
ea
ch
m
eth
o
d
is
illu
s
tr
ated
in
Fi
g
u
r
e
9
.
Fi
g
u
r
e
9
.
T
h
e
C
o
m
p
u
tatio
n
T
im
e
o
f
Face
R
ec
o
g
n
itio
n
(
m
illi
s
ec
o
n
d
)
4.
CO
NCLU
SI
O
N
Gen
er
all
y
,
t
h
e
b
ig
p
r
o
b
lem
i
n
f
ac
e
r
ec
o
g
n
it
io
n
s
y
s
te
m
is
h
o
w
to
s
o
lv
e
t
h
e
v
ar
iatio
n
o
f
p
o
s
es
an
d
th
e
v
ar
iatio
n
o
f
ill
u
m
in
a
n
ce
to
g
et
th
e
b
est
r
ec
o
g
n
itio
n
r
ate
an
d
t
o
d
ec
r
ea
s
e
r
ec
o
g
n
itio
n
ti
m
e.
I
n
t
h
is
s
tu
d
y
w
e
u
s
e
t
w
o
le
n
s
es
o
n
s
ter
eo
v
is
io
n
ca
m
er
a.
Gen
er
all
y
,
a
s
ter
eo
v
i
s
i
o
n
ca
m
er
a
is
u
s
ed
i
n
a
f
ac
e
r
ec
o
g
n
itio
n
s
y
s
te
m
to
r
ec
o
n
s
tr
u
ct
d
o
ts
o
f
f
ac
e
t
h
en
t
o
p
r
o
d
u
ce
a
th
r
ee
-
d
i
m
e
n
s
io
n
a
l
(
3
D)
f
ac
e
i
m
a
g
e.
Ho
w
e
v
er
,
th
e
p
r
o
ce
s
s
o
f
f
ac
e
r
ec
o
g
n
itio
n
u
s
i
n
g
3
D
i
m
a
g
es
r
eq
u
ir
e
m
o
r
e
r
ec
o
g
n
it
io
n
ti
m
e
t
h
an
th
e
u
s
e
o
f
t
w
o
-
d
i
m
en
s
io
n
al
(
2
D)
f
ac
e
i
m
a
g
es.
T
h
is
r
esear
c
h
p
r
o
p
o
s
es
a
m
o
d
el
o
f
f
ac
e
r
ec
o
g
n
itio
n
u
s
i
n
g
t
h
e
m
et
h
o
d
o
f
j
o
in
in
g
t
w
o
f
ac
e
i
m
a
g
e
s
f
r
o
m
le
f
t
an
d
r
i
g
h
t
le
n
s
f
r
o
m
a
s
ter
eo
v
i
s
io
n
ca
m
er
a
n
a
m
el
y
as
y
m
m
etr
ical
h
al
f
-
j
o
in
m
e
t
h
o
d
.
As
y
m
m
etr
ical
h
al
f
-
j
o
in
m
et
h
o
d
is
u
s
ed
i
n
f
a
ce
i
m
a
g
e
n
o
r
m
aliza
tio
n
p
r
o
ce
s
s
i
n
g
.
I
n
as
y
m
m
etr
ical
h
alf
-
j
o
in
m
et
h
o
d
,
a
R
o
I
o
f
f
ac
e
i
m
a
g
e
f
r
o
m
le
f
t
an
d
r
ig
h
t
len
s
es
ar
e
p
r
o
d
u
ce
d
b
ased
o
n
ce
n
ter
ax
is
o
f
ea
ch
e
y
e
f
r
o
m
e
y
e
d
ete
ctio
n
.
T
h
e
cr
o
p
p
in
g
f
ac
e
i
m
a
g
e
f
r
o
m
as
y
m
m
etr
ica
l
h
al
f
-
j
o
in
m
o
d
el
h
as
d
i
f
f
er
en
t
w
id
e
o
f
cr
o
p
p
in
g
d
ep
en
d
o
n
e
y
e
s
co
o
r
d
in
at
e
lo
ca
tio
n
.
B
ased
o
n
th
e
e
x
p
er
i
m
e
n
ts
,
t
h
e
p
r
o
p
o
s
ed
s
y
s
te
m
s
h
o
w
s
t
h
at
t
h
e
as
y
m
m
e
tr
ical
h
a
lf
-
j
o
in
m
et
h
o
d
is
ab
le
to
p
r
o
d
u
ce
av
er
ag
e
9
7
.
5
%
o
f
f
ac
e
r
ec
o
g
n
itio
n
r
ate
a
n
d
h
a
s
an
a
v
er
ag
e
co
m
p
u
tat
io
n
ti
m
e
o
f
f
ac
e
r
ec
o
g
n
itio
n
o
f
4
5
.
9
9
m
s
.
T
h
e
as
y
m
m
etr
ic
al
h
al
f
-
j
o
in
m
et
h
o
d
h
as
a
b
et
ter
ac
cu
r
ac
y
t
h
an
f
ac
e
r
ec
o
g
n
itio
n
s
y
s
te
m
u
s
i
n
g
s
in
g
le
v
is
io
n
m
et
h
o
d
an
d
s
y
m
m
etr
ical
h
a
lf
-
j
o
in
m
et
h
o
d
.
T
h
is
s
tu
d
y
s
h
o
w
s
th
a
t
th
e
s
ter
eo
v
is
io
n
ca
m
er
a
h
a
s
a
ca
p
ab
ilit
y
to
p
r
o
d
u
ce
t
w
o
o
f
2
D
f
ac
e
i
m
ag
e
s
f
r
o
m
each
le
n
s
th
at
is
ab
le
to
b
e
elab
o
r
ated
w
it
h
2D
f
ac
e
i
m
a
g
e
r
ec
o
g
n
itio
n
tech
n
iq
u
e
a
n
d
also
to
p
r
o
d
u
ce
h
ig
h
r
ec
o
g
n
itio
n
r
ate
an
d
s
h
o
r
t
r
ec
o
g
n
itio
n
ti
m
e
.
RE
F
E
R
E
NC
E
S
[1
]
G
.
P
.
Ku
su
m
a
a
n
d
C.
S
.
Ch
u
a
,
“
P
CA
-
b
a
se
d
ima
g
e
re
c
o
m
b
in
a
ti
o
n
f
o
r
m
u
lt
i
m
o
d
a
l
2
D+
3
D
f
a
c
e
re
c
o
g
n
it
io
n
,
”
Ima
g
e
a
n
d
Vi
si
o
n
C
o
mp
u
ti
n
g
,
v
o
l/
issu
e
:
29
(
5
),
pp.
3
0
6
-
3
1
6
,
2
0
1
1
.
[2
]
A.
M
o
e
in
i,
e
t
a
l
.,
“
Ex
p
re
ss
io
n
-
I
n
v
a
rian
t
F
a
c
e
Re
c
o
g
n
it
io
n
v
ia
3
D
F
a
c
e
Re
c
o
n
stru
c
ti
o
n
Us
in
g
G
a
b
o
r
F
i
l
ter
Ba
n
k
f
ro
m
a
2
D
S
in
g
le
Im
a
g
e
,”
i
n
Pa
tt
e
rn
Rec
o
g
n
i
ti
o
n
(
ICPR
),
2
0
1
4
2
2
n
d
I
n
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
,
p
p
.
4
7
0
8
-
4
7
1
3
,
2
0
1
4
.
[3
]
H.
M
.
Ra
ra
,
e
t
a
l
.,
“
Dista
n
t
f
a
c
e
re
c
o
g
n
it
io
n
b
a
se
d
o
n
s
p
a
rse
-
ste
re
o
re
c
o
n
stru
c
ti
o
n
,”
i
n
Im
a
g
e
Pr
o
c
e
ss
in
g
(
ICIP)
,
2
0
0
9
1
6
t
h
IE
EE
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
,
p
p
.
4
1
4
1
-
4
1
4
4
,
2
0
0
9
.
[4
]
E.
T
e
m
iz
k
a
n
a
n
d
H.
S
.
Bil
g
e
, “
S
h
a
p
e
in
d
e
x
b
a
se
d
3
d
f
a
c
e
r
e
c
o
g
n
it
io
n
u
sin
g
p
h
o
t
o
m
e
tri
c
ste
re
o
,”
i
n
2
0
1
1
IEE
E
1
9
t
h
S
ig
n
a
l
Pro
c
e
ss
in
g
a
n
d
C
o
mm
u
n
ica
ti
o
n
s A
p
p
li
c
a
ti
o
n
s Co
n
fer
e
n
c
e
(S
IU)
,
p
p
.
4
5
4
-
4
5
7
,
2
0
1
1
.
[5
]
M
.
A
.
A
b
d
e
lrah
m
a
n
,
e
t
a
l
.,
“
A
p
a
ss
iv
e
ste
r
e
o
s
y
ste
m
f
o
r
3
D
h
u
m
a
n
f
a
c
e
re
c
o
n
stru
c
ti
o
n
a
n
d
re
c
o
g
n
it
io
n
a
t
a
d
istan
c
e
,”
i
n
2
0
1
2
IEE
E
C
o
m
p
u
ter
S
o
c
iety
Co
n
fer
e
n
c
e
o
n
Co
mp
u
ter
Vi
sio
n
a
n
d
Pa
tt
e
rn
Rec
o
g
n
it
i
o
n
W
o
rk
sh
o
p
s
,
p
p
.
1
7
-
2
2
,
2
0
1
2
.
[6
]
C.
Ho
lzm
a
n
n
a
n
d
M.
Ho
c
h
g
a
tt
e
re
r
,
“
M
e
a
su
rin
g
d
istan
c
e
w
it
h
m
o
b
il
e
p
h
o
n
e
s
u
si
n
g
sin
g
le
-
c
a
m
e
ra
ste
re
o
v
isio
n
,
”
D
istrib
u
ted
Co
mp
u
ti
n
g
S
y
ste
ms
W
o
rk
sh
o
p
s (
ICDCS
W
),
3
2
n
d
In
te
rn
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
,
IEE
E,
p
p
.
88
-
9
3
,
2
0
1
2
.
[7
]
E.
W
in
a
rn
o
,
e
t
a
l
.
,
“
De
v
e
lo
p
m
e
n
t
o
f
F
a
c
e
R
e
c
o
g
n
it
io
n
S
y
ste
m
a
n
d
F
a
c
e
Dista
n
c
e
Esti
m
a
t
io
n
Us
in
g
S
tere
o
V
isio
n
Ca
m
e
ra
,
”
J
o
u
rn
a
l
o
f
T
h
e
o
re
ti
c
a
l
a
n
d
Ap
p
li
e
d
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
/i
ss
u
e
:
67
(
3
)
,
p
p
.
6
5
2
-
6
5
7
,
2
0
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
2
0
1
7
:
3
4
1
1
–
3
4
2
0
3420
[8
]
J.
Ha
rg
u
e
ss
a
n
d
J.
K.
Ag
g
a
r
wa
l,
“
A
c
a
se
f
o
r
th
e
a
v
e
ra
g
e
-
h
a
lf
-
f
a
c
e
in
2
D
a
n
d
3
D
f
o
r
f
a
c
e
re
c
o
g
n
it
io
n
,”
i
n
C
o
mp
u
ter
Vi
sio
n
a
n
d
Pa
tt
e
rn
Rec
o
g
n
it
io
n
W
o
rk
sh
o
p
s,
2
0
0
9
.
CVP
R
W
o
rk
sh
o
p
s
2
0
0
9
.
IEE
E
Co
m
p
u
ter
S
o
c
iet
y
Co
n
fer
e
n
c
e
o
n
,
p
p
.
7
-
12
,
2
0
0
9
.
[9
]
S.
Zaf
e
iri
o
u
,
e
t
a
l.
,
“
F
a
c
e
re
c
o
g
n
it
io
n
a
n
d
v
e
rif
ica
ti
o
n
u
sin
g
p
h
o
to
m
e
tri
c
ste
re
o
:
T
h
e
p
h
o
to
f
a
c
e
d
a
tab
a
se
a
n
d
a
c
o
m
p
re
h
e
n
siv
e
e
v
a
lu
a
ti
o
n
,”
In
f
o
rm
a
ti
o
n
Fo
re
n
sic
s
a
n
d
S
e
c
u
rity,
I
EE
E
T
ra
n
sa
c
ti
o
n
s
o
n
,
v
o
l/
iss
u
e
:
8
(1
),
p
p
.
1
2
1
-
135
,
2
0
1
3
.
[1
0
]
E.
W
in
a
rn
o
,
e
t
a
l
.
,
“
Im
p
ro
v
e
d
Re
a
l
-
T
i
m
e
F
a
c
e
R
e
c
o
g
n
it
io
n
Ba
se
d
On
T
h
re
e
Lev
e
l
W
a
v
e
let
De
c
o
m
p
o
siti
o
n
-
P
ri
n
c
ip
a
l
Co
m
p
o
n
e
n
t
A
n
a
ly
sis
A
n
d
M
a
h
a
lan
o
b
is
Dista
n
c
e
,
”
J
o
u
rn
a
l
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
v
o
l/
issu
e
:
10
(
5
)
,
p
p
.
844
-
8
5
1
,
2
0
1
4
.
[1
1
]
D.
S
rid
h
a
r
a
n
d
I.
M
.
Krish
n
a
,
“
F
a
c
e
Re
c
o
g
n
it
io
n
Us
in
g
T
w
o
D
im
e
n
sio
n
a
l
Disc
re
te
Co
sin
e
T
ra
n
sf
o
rm
,
L
in
e
a
r
Disc
ri
m
in
a
n
t
A
n
a
ly
sis
a
n
d
K
Ne
a
re
st
Ne
ig
h
b
o
r
Clas
sif
ier
,
”
IAE
S
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Art
if
ici
a
l
I
n
telli
g
e
n
c
e
,
v
o
l/
issu
e
:
1
(
4
),
pp.
1
6
1
,
2
0
1
2
.
[1
2
]
S.
N.
Ra
n
d
iv
e
a
n
d
A
.
B.
G
o
n
d
e
,
“
A
n
o
v
e
l
a
p
p
ro
a
c
h
f
o
r
fa
c
e
re
c
o
g
n
it
io
n
u
sin
g
f
u
sio
n
o
f
lo
c
a
l
g
a
b
o
r
p
a
tt
e
rn
s
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
Co
mp
u
ter
E
n
g
i
n
e
e
rin
g
(
IJ
ECE
)
,
v
o
l/
issu
e
:
2
(3
)
,
pp.
3
4
5
,
2
0
1
6
.
[1
3
]
E.
W
in
a
rn
o
,
e
t
a
l.
“
F
a
c
e
Re
c
o
g
n
it
io
n
Ba
se
d
o
n
S
y
m
m
e
t
rica
l
Ha
lf
-
Jo
in
M
e
th
o
d
u
sin
g
S
tere
o
V
isi
o
n
Ca
m
e
r
a
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
Co
mp
u
ter
E
n
g
i
n
e
e
rin
g
,
v
o
l/
issu
e
:
6
(
6
)
,
p
p
.
2
8
1
8
-
2
8
2
7
,
2
0
1
6
.
[1
4
]
P.
V
i
o
la
a
n
d
M.
Jo
n
e
s
,
“
Ra
p
id
o
b
jec
t
d
e
tec
ti
o
n
u
sin
g
a
b
o
o
ste
d
c
a
sc
a
d
e
o
f
si
m
p
le
f
e
a
tu
re
s
,
”
i
n
Pr
o
c
e
e
d
in
g
IE
E
E
Co
n
fer
e
n
c
e
o
n
c
o
mp
u
ter
v
isio
n
a
n
d
p
a
t
ter
n
re
c
o
g
n
it
i
o
n
2
0
0
1
,
p
p
.
1
-
9
,
2
0
0
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.