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.
9
,
No
.
6
,
Dec
em
b
er
201
9
,
p
p
.
4
7
0
3
~4
7
1
2
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v9
i
6
.
p
p
4
7
0
3
-
4712
4703
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
Dua
l iris auth
enti
ca
tion sy
ste
m
usi
ng
de
z
ert
s
m
a
ra
nda
che
the
o
ry
K
a
m
el
G
ha
ne
m
G
ha
le
m
1
,
F
a
t
iha
H
endel
2
1
Eco
le
S
u
p
é
rieu
re
E
n
G
é
n
ie E
lec
tri
q
u
e
e
t
E
n
e
rg
é
ti
q
u
e
,
A
lg
é
rie
2
Un
iv
e
rsité d
e
s S
c
ien
c
e
s e
t
d
e
la
T
e
c
h
n
o
lo
g
ie d
’Ora
n
M
o
h
a
m
e
d
Bo
u
d
iaf
,
A
l
g
é
rie
Art
icle
I
n
fo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
2
5
,
2
0
1
8
R
ev
i
s
ed
Ma
y
16
,
2
0
1
9
A
cc
ep
ted
J
u
n
2
6
,
2
0
1
9
In
th
is
p
a
p
e
r,
a
d
u
a
l
iri
s
a
u
th
e
n
ti
c
a
ti
o
n
u
si
n
g
De
z
e
rt
S
m
a
ra
n
d
a
c
h
e
th
e
o
ry
is
p
re
se
n
ted
.
T
h
e
p
r
o
p
o
se
d
m
e
th
o
d
c
o
n
sists
o
f
th
re
e
m
a
in
ste
p
s:
In
th
e
f
irst
o
n
e
,
th
e
iri
s
im
a
g
e
s
a
re
se
g
m
e
n
ted
in
o
rd
e
r
t
o
e
x
trac
t
o
n
ly
h
a
lf
iri
s
d
isc
t
h
a
t
c
o
n
tain
s
re
lev
a
n
t
i
n
f
o
rm
a
ti
o
n
a
n
d
is
les
s
a
ff
e
c
ted
b
y
n
o
ise
.
F
o
r
th
a
t
,
a
Ho
u
g
h
tran
sf
o
r
m
is u
se
d
.
T
h
e
se
g
m
e
n
ted
ima
g
e
s a
r
e
n
o
rm
a
li
z
e
d
b
y
Da
u
g
m
a
n
ru
b
b
e
r
sh
e
e
t
m
o
d
e
l.
In
th
e
se
c
o
n
d
ste
p
,
th
e
n
o
rm
a
li
z
e
d
i
m
a
g
e
s
a
r
e
a
n
a
l
y
z
e
d
b
y
a
b
e
n
c
h
o
f
tw
o
1
D
L
o
g
-
G
a
b
o
r
f
i
lt
e
rs
to
e
x
tra
c
t
th
e
tex
tu
re
c
h
a
r
a
c
teristics
.
T
h
e
e
n
c
o
d
in
g
is
re
a
li
z
e
d
w
it
h
a
p
h
a
se
o
f
q
u
a
n
ti
z
a
ti
o
n
d
e
v
e
lo
p
e
d
b
y
J.
Da
u
g
m
a
n
to
g
e
n
e
ra
te
th
e
b
in
a
ry
iri
s
tem
p
late
.
F
o
r
th
e
a
u
t
h
e
n
ti
c
a
ti
o
n
a
n
d
t
h
e
sim
il
a
rit
y
m
e
a
su
re
m
e
n
t
b
e
tw
e
e
n
b
o
t
h
b
in
a
ry
iri
se
s
te
m
p
late
s,
th
e
h
a
m
m
in
g
d
istan
c
e
s
a
re
u
se
d
w
it
h
a
p
re
v
io
u
sl
y
c
a
lcu
late
d
th
re
sh
o
ld
.
T
h
e
sc
o
re
f
u
sio
n
is
a
p
p
li
e
d
u
sin
g
DSm
C
c
o
m
b
in
a
ti
o
n
ru
le.
T
h
e
p
ro
p
o
se
d
m
e
th
o
d
h
a
s
b
e
e
n
tes
ted
o
n
a
su
b
se
t
o
f
iri
s
d
a
ta
b
a
se
CA
S
I
A
-
IrisV
3
-
In
terv
a
l.
T
h
e
o
b
tain
e
d
r
e
su
lt
s
g
iv
e
a
sa
ti
s
f
a
c
to
r
y
p
e
r
f
o
r
m
a
n
c
e
w
it
h
a
c
c
u
ra
c
y
o
f
9
9
.
9
6
%
,
F
A
R
o
f
0
%
,
F
RR
o
f
3
.
8
9
%
,
EE
R
o
f
2
%
a
n
d
p
ro
c
e
ss
in
g
ti
m
e
f
o
r
o
n
e
iri
s im
a
g
e
o
f
1
2
.
3
6
s.
K
ey
w
o
r
d
s
:
Au
t
h
en
t
icatio
n
B
io
m
etr
ic
Dez
er
t S
m
ar
an
d
ac
h
e
th
eo
r
y
I
r
is
Co
p
y
rig
h
t
©
2
0
1
9
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
:
Ka
m
el
G
h
an
e
m
G
h
ale
m
,
E
co
le
Su
p
ér
ieu
r
e
E
n
Gé
n
ie
E
l
ec
tr
iq
u
e
et
E
n
er
g
étiq
u
e
,
B
P
6
4
C
H2
A
C
H
A
B
A
H
ANI
FI
UST
O
3
1
0
0
0
Or
an
,
A
lg
ér
ie
.
E
m
ail:
g
h
a
l
e
m
_
k
a
m
e
l_
g
h
an
e
m
@
es
g
ee
-
o
r
an
.
d
z
1.
I
NT
RO
D
UCT
I
O
N
W
h
en
in
d
i
v
id
u
al
s
lo
g
o
n
to
co
m
p
u
ter
s
,
o
r
ac
ce
s
s
an
A
T
M,
o
r
p
ass
th
r
o
u
g
h
air
p
o
r
t
s
ec
u
r
it
y
,
th
e
y
h
a
v
e
to
r
ev
ea
l
th
eir
id
en
titi
e
s
.
Fo
r
th
is
,
i
n
d
iv
id
u
als
u
s
e
p
ass
w
o
r
d
s
,
A
T
M
ca
r
d
s
,
an
d
p
ass
ep
o
r
ts
to
p
r
o
v
e
th
eir
id
en
titi
e
s
.
Ho
w
e
v
er
,
p
ass
w
o
r
d
s
ca
n
b
e
f
o
r
g
o
tten
,
a
n
d
AT
M
ca
r
d
s
o
r
p
ass
p
o
r
ts
ca
n
b
e
lo
s
t
o
r
s
to
len
.
I
n
co
n
tr
ar
y
,
t
h
e
b
io
m
etr
ic
m
o
d
alities
(
Fin
g
er
p
r
in
t,
f
ac
e,
ir
is
,
…etc)
s
p
ea
k
to
w
h
at
an
d
th
e
y
al
s
o
allo
w
to
p
r
o
v
e
o
u
r
id
en
tit
y
.
Ho
w
e
v
er
,
th
e
u
n
i
m
o
d
al
b
io
m
etr
ic
s
y
s
te
m
s
u
s
i
n
g
o
n
e
b
io
m
e
tr
ic
m
o
d
alit
y
f
o
r
r
ec
o
g
n
i
tio
n
ca
n
n
o
t
g
u
ar
a
n
tee
at
p
r
ese
n
t
a
n
e
x
ce
ll
en
t
r
ec
o
g
n
itio
n
r
ate.
F
u
r
th
er
m
o
r
e,
th
ese
s
y
s
te
m
s
s
u
f
f
er
f
r
o
m
li
m
i
tatio
n
s
s
u
c
h
as
s
en
s
iti
v
it
y
to
n
o
is
e,
d
ata
q
u
alit
y
,
n
o
n
-
u
n
i
v
er
s
al
it
y
,
an
d
s
p
o
o
f
attac
k
s
.
T
o
o
v
er
co
m
e
th
e
s
e
p
r
o
b
lem
s
,
Mu
lti
m
o
d
al
b
io
m
e
tr
ic
s
y
s
te
m
s
,
w
h
ich
co
m
b
i
n
e
m
u
ltip
le
b
io
m
etr
ic
m
o
d
alitie
s
,
h
av
e
b
ee
n
d
ev
elo
p
ed
o
n
p
u
r
p
o
s
e
to
ac
h
iev
e
a
b
etter
r
ec
o
g
n
itio
n
r
ate.
T
h
e
p
o
p
u
lar
f
u
s
io
n
m
et
h
o
d
o
f
th
e
b
io
m
etr
ic
tr
aits
ca
n
b
e
d
o
n
e
at
to
w
s
ta
g
es o
f
r
ec
o
g
n
it
io
n
s
y
s
t
e
m
:
1.
Fu
s
io
n
at
f
ea
tu
r
e
e
x
tr
ac
tio
n
le
v
el
T
h
e
d
ata
is
ac
q
u
ir
ed
f
r
o
m
ea
ch
s
en
s
o
r
is
u
tili
ze
d
to
g
en
er
ate
a
f
ea
tu
r
e
v
ec
to
r
.
T
h
en
,
th
e
f
ea
tu
r
es
ar
e
f
u
s
ed
to
f
o
r
m
o
n
e
f
ea
t
u
r
e
v
ec
t
o
r
.
2.
Fu
s
io
n
at
m
atch
in
g
s
co
r
e
lev
el
T
h
e
m
atc
h
i
n
g
s
co
r
e
o
f
ea
ch
s
y
s
te
m
i
s
co
m
b
i
n
ed
a
n
d
co
m
p
ar
ed
w
it
h
th
e
s
to
r
ed
te
m
p
late.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
6
,
Dec
em
b
er
2
0
1
9
:
4
7
0
3
-
4
7
1
2
4704
W
e
u
s
e
as
m
o
d
alit
y
f
o
r
r
ec
o
g
n
itio
n
o
f
i
n
d
iv
id
u
als:
ir
i
s
,
s
i
n
ce
th
eir
tex
tu
r
e
is
-
Stab
le
t
h
r
o
u
g
h
o
u
t t
h
e
li
f
e
o
f
a
p
er
s
o
n
,
u
n
li
k
e
t
h
e
f
i
n
g
er
p
r
in
t.
-
Un
iq
u
e
f
o
r
ea
ch
p
er
s
o
n
,
u
n
li
k
e
a
f
ac
ial
f
ea
t
u
r
e
in
id
en
tical
t
w
i
n
s
.
-
Un
f
al
s
i
f
iab
le
co
n
tr
ar
y
to
th
e
ch
ar
ac
ter
is
tic
s
o
f
t
h
e
v
o
ice.
-
I
r
is
is
an
i
n
ter
n
a
l
o
r
g
an
w
ell
p
r
o
tecte
d
f
r
o
m
t
h
e
ex
t
er
n
al
en
v
ir
o
n
m
en
t,
b
u
t
n
e
v
e
r
th
eles
s
m
ea
s
u
r
ab
le,
in
a
r
ath
er
litt
le
i
n
v
asi
v
e
w
a
y
,
b
y
a
s
i
m
p
le
i
m
a
g
e
ac
q
u
is
itio
n
.
Dau
g
m
a
n
’
s
al
g
o
r
ith
m
[
1
]
i
s
o
n
e
o
f
th
e
b
est
ir
is
alg
o
r
it
h
m
k
n
o
w
n
i
n
b
io
m
etr
ics.
T
h
e
alg
o
r
ith
m
co
n
s
is
ts
to
s
e
g
m
e
n
t
ir
i
s
u
s
in
g
I
n
te
g
r
o
-
Di
f
f
er
en
tia
l
Op
er
ato
r
an
d
ir
is
n
o
r
m
aliza
tio
n
is
i
m
p
le
m
en
ted
u
s
i
n
g
Dau
g
m
a
n
’
s
p
o
lar
r
ep
r
esen
tati
o
n
.
T
h
en
,
ir
is
e
n
co
d
in
g
is
ap
p
lied
u
s
i
n
g
2
D
Gab
o
r
f
ilte
r
s
to
ex
tr
ac
t
a
b
in
ar
y
co
d
e
o
f
2
5
6
b
y
tes.
T
h
e
Ma
tch
i
n
g
i
s
p
r
o
ce
s
s
ed
b
y
co
m
p
u
tin
g
s
i
m
ilar
i
t
y
b
et
w
ee
n
t
wo
ir
is
co
d
es
u
s
in
g
Ha
m
m
i
n
g
d
i
s
ta
n
ce
.
T
h
e
m
o
r
e
Ha
m
m
i
n
g
d
is
ta
n
ce
is
s
m
all,
t
h
e
m
o
r
e
b
o
th
co
d
es
ar
e
s
i
m
il
ar
.
A
d
is
ta
n
ce
o
f
0
co
r
r
esp
o
n
d
s
to
a
p
er
f
ec
t
m
atch
b
et
w
ee
n
b
o
th
ir
is
i
m
a
g
es,
w
h
i
le
t
w
o
ir
i
s
i
m
ag
e
s
o
f
d
i
f
f
er
e
n
t
p
er
s
o
n
w
ill
h
av
e
a
Ha
m
m
i
n
g
d
is
tan
ce
clo
s
e
to
0
.
5
0
.
I
n
1
9
9
7
,
W
ild
es
[
2
]
p
r
o
p
o
s
ed
a
n
o
v
e
l
ir
is
r
ec
o
g
n
i
tio
n
s
y
s
te
m
co
m
p
ar
ed
to
Da
u
g
m
a
n
al
g
o
r
ith
m
[
1
]
.
T
h
e
ac
q
u
is
itio
n
o
f
ir
i
s
is
d
o
n
e
b
y
a
C
C
D
C
a
m
er
a
in
lo
w
l
u
m
in
o
s
it
y
.
T
h
en
,
t
h
e
ir
is
is
s
eg
m
e
n
ted
u
s
in
g
C
ir
cu
lar
an
d
E
llip
tic
Ho
u
g
h
tr
an
s
f
o
r
m
an
d
is
n
o
r
m
alize
d
u
s
i
n
g
a
tr
an
s
f
o
r
m
atio
n
f
u
n
ctio
n
o
f
p
ix
els.
Af
ter
th
at,
th
e
ir
i
s
is
f
ilter
ed
b
y
L
ap
lac
ian
o
f
Ga
u
s
s
ian
f
i
lter
s
w
i
th
f
o
u
r
d
i
f
f
er
en
t
r
e
s
o
lu
tio
n
le
v
els
.
A
n
o
r
m
alize
d
co
r
r
elatio
n
is
ca
lc
u
lated
f
o
r
e
v
er
y
r
e
s
o
lu
tio
n
le
v
el
s
.
T
h
e
m
e
d
ian
o
f
th
e
v
a
lu
e
s
o
f
co
r
r
elatio
n
s
i
s
co
m
p
u
ted
f
o
r
th
e
f
ilter
ed
i
m
a
g
e.
T
h
e
f
u
s
io
n
o
f
f
o
u
r
v
alu
e
s
is
ap
p
lied
u
s
i
n
g
Fis
h
er
's lin
ea
r
d
is
cr
i
m
i
n
an
t.
I
n
1
9
9
8
,
W
.
B
o
les
an
d
B
.
B
o
ash
a
s
h
[
3
]
,
p
r
esen
ted
a
n
e
w
a
lg
o
r
ith
m
f
o
r
r
ec
o
g
n
itio
n
o
f
i
n
d
iv
id
u
al
s
f
r
o
m
ir
is
i
m
ag
e
s
.
T
h
e
alg
o
r
ith
m
i
s
in
s
e
n
s
iti
v
e
to
v
ar
iati
o
n
in
th
e
li
g
h
ti
n
g
co
n
d
itio
n
s
an
d
n
o
is
e
lev
el
s
.
A
Me
d
ia
n
f
ilter
is
u
s
ed
f
o
r
p
r
ep
r
o
ce
s
s
in
g
.
T
h
e
ad
v
a
n
ta
g
e
o
f
th
is
tech
n
iq
u
e
is
to
e
x
tr
ac
t
a
f
ea
tu
r
es
v
ec
to
r
f
r
o
m
1
D
s
ig
n
al
s
r
ath
er
th
a
n
2
D
i
m
a
g
es
a
n
al
y
ze
d
i
n
[
1
]
,
[
2
]
u
s
in
g
ze
r
o
-
cr
o
s
s
in
g
s
o
f
t
h
e
d
y
ad
ic
w
a
v
elet
tr
an
s
f
o
r
m
at
v
ar
io
u
s
r
eso
lu
t
io
n
le
v
els.
On
l
y
a
f
e
w
s
elec
ted
i
n
ter
m
ed
iate
r
eso
lu
tio
n
s
ar
e
u
s
ed
f
o
r
m
atc
h
in
g
.
T
h
e
m
atc
h
in
g
is
ap
p
lied
u
s
i
n
g
d
i
f
f
er
en
t
d
is
s
i
m
ilar
it
y
f
u
n
ctio
n
s
.
T
h
u
s
,
m
a
k
e
th
e
a
lg
o
r
it
h
m
f
a
s
ter
an
d
le
s
s
s
e
n
s
i
tiv
e
to
n
o
is
e
an
d
q
u
an
ti
f
ica
tio
n
er
r
o
r
.
I
n
2
0
0
4
,
Ma
et
al.
[
4
]
p
r
esen
ted
an
ef
f
ic
ien
t
al
g
o
r
ith
m
f
o
r
ir
is
r
ec
o
g
n
itio
n
.
T
h
e
ir
is
is
s
eg
m
en
ted
b
y
C
an
n
y
f
ilter
a
n
d
Ho
u
g
h
tr
a
n
s
f
o
r
m
.
T
h
e
n
,
t
h
e
ir
is
is
n
o
r
m
alize
d
b
y
h
i
s
to
g
r
a
m
eq
u
al
izatio
n
.
Af
ter
t
h
at,
A
1
D
W
av
elet
T
r
an
s
f
o
r
m
is
u
s
ed
to
r
ep
r
esen
t
r
es
u
lti
n
g
1
D
in
te
n
s
it
y
s
i
g
n
als.
T
h
e
p
o
s
itio
n
o
f
lo
ca
l
s
h
ar
p
v
ar
iatio
n
p
o
i
n
ts
is
r
e
g
is
ter
e
d
as
f
ea
t
u
r
es.
T
h
e
m
atc
h
i
n
g
is
ef
f
ec
t
u
ated
u
s
i
n
g
t
h
e
s
i
m
ilar
it
y
f
u
n
ctio
n
(
ex
clu
s
iv
e
o
r
o
p
er
atio
n
)
.
T
h
is
alg
o
r
ith
m
is
e
f
f
icie
n
t a
n
d
f
aste
r
th
an
Da
u
g
m
a
n
’
s
al
g
o
r
ith
m
[
1
]
.
I
n
[
5
]
,
th
e
r
esear
ch
er
s
p
r
o
p
o
s
ed
a
m
o
d
if
ied
Ma
s
e
k
a
p
p
r
o
ac
h
an
d
a
co
m
p
ar
ati
v
e
s
t
u
d
y
o
f
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
f
o
llo
w
i
n
g
m
et
h
o
d
s
:
r
ad
ial
s
eg
m
e
n
tatio
n
,
Ma
s
ek
s
e
g
m
en
ta
tio
n
a
p
p
r
o
ac
h
,
m
o
d
i
f
ied
Ma
s
ek
ap
p
r
o
ac
h
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
test
ed
o
n
C
asia
I
r
is
Data
b
ase
V3
s
h
o
w
ed
a
g
o
o
d
p
e
r
f
o
r
m
a
n
ce
i
n
ter
m
s
o
f
ac
c
u
r
ac
y
a
n
d
p
r
o
ce
s
s
in
g
t
i
m
e.
R
.
B
is
w
as
[
6
]
h
a
s
i
n
tr
o
d
u
ce
d
an
ir
is
r
ec
o
g
n
it
io
n
s
y
s
te
m
th
a
t
in
cl
u
d
es
d
i
f
f
er
e
n
t
s
tep
s
:
s
e
g
m
en
tatio
n
,
n
o
r
m
aliza
t
io
n
,
f
ea
t
u
r
e
ex
tr
ac
tio
n
,
an
d
class
if
ica
tio
n
.
T
h
e
s
eg
m
en
tati
o
n
o
f
t
h
e
p
u
p
il
is
p
er
f
o
r
m
ed
b
y
th
e
Ho
u
g
h
tr
an
s
f
o
r
m
.
E
x
p
er
i
m
en
tal
r
es
u
lts
s
h
o
w
ed
a
r
ec
o
g
n
i
tio
n
r
ate
o
f
9
2
%.
I
n
[
7
]
,
th
e
a
u
t
h
o
r
s
p
r
o
p
o
s
ed
an
ir
is
r
ec
o
g
n
itio
n
s
y
s
te
m
b
ase
d
o
n
"
Fra
ctal
d
i
m
e
n
s
io
n
o
f
b
o
x
-
co
u
n
tin
g
m
et
h
o
d
"
.
First,
th
e
ir
is
is
s
e
g
m
e
n
ted
b
y
Ho
u
g
h
tr
an
s
f
o
r
m
an
d
is
n
o
r
m
alize
d
b
y
Da
u
g
m
a
n
’
s
r
u
b
b
er
s
h
ee
t
m
o
d
el.
T
h
e
n
,
t
h
e
f
ea
tu
r
e
e
x
tr
ac
tio
n
is
p
r
o
ce
s
s
ed
b
y
b
o
x
co
u
n
t
in
g
.
F
in
al
l
y
,
t
h
e
m
atc
h
i
n
g
is
estab
li
s
h
ed
u
s
in
g
K
-
n
ea
r
est
Nei
g
h
b
o
r
an
d
E
u
cl
id
ea
n
d
is
tan
ce
.
E
x
p
er
i
m
e
n
ts
t
ested
o
n
C
asia
I
n
ter
v
al
V4
d
atab
ase
s
h
o
w
ed
a
g
o
o
d
r
ec
o
g
n
itio
n
r
ate
eq
u
al
to
9
2
.
6
3
%.
D.
B
o
b
eld
y
k
a
n
d
A
.
R
o
s
s
[
8
]
h
av
e
d
ev
elo
p
ed
a
m
et
h
o
d
f
o
r
p
r
e
d
ictin
g
e
y
e
co
lo
r
f
r
o
m
NI
R
ir
is
i
m
a
g
es.
R
e
s
ea
r
ch
er
s
h
av
e
s
h
o
w
n
th
a
t
a
tex
t
u
r
e
b
ased
ap
p
r
o
ac
h
b
ased
o
n
th
e
B
SIF
is
m
o
r
e
ef
f
ic
ien
t
t
h
a
n
th
e
in
te
n
s
it
y
b
ased
ap
p
r
o
ac
h
b
ased
o
n
r
a
w
p
ix
el
v
al
u
es.
E
x
p
er
im
en
ts
te
s
ted
o
n
t
h
e
B
io
C
OP
d
atab
ase
s
h
o
w
ed
a
g
o
o
d
r
ec
o
g
n
itio
n
r
ate
o
f
9
0
%.
T
h
e
B
SIF
d
is
tin
g
u
is
h
ed
"
lig
h
t
co
lo
r
ir
is
"
an
d
"
d
ar
k
co
lo
r
ir
id
es
"
u
s
in
g
th
e
SVM
cla
s
s
i
f
ier
.
T
h
e
au
th
o
r
s
[
9
]
p
r
esen
ted
a
n
ew
m
eth
o
d
o
f
class
i
f
y
i
n
g
f
a
k
ed
ir
is
i
m
ag
e
s
o
f
d
if
f
er
e
n
t
p
atter
n
s
s
u
ch
as
p
r
in
ted
ir
is
es,
co
n
tact
le
n
s
es.
T
h
e
n
e
w
clas
s
i
f
icatio
n
m
e
th
o
d
lear
n
s
d
if
f
er
en
t
ch
ar
ac
ter
i
s
tics
o
f
f
ak
ed
ir
i
s
i
m
a
g
es
b
y
C
NN
an
d
id
en
ti
f
ie
s
leg
iti
m
ate
a
n
d
f
a
k
ed
ir
is
i
m
ag
es
u
s
i
n
g
"
Hier
ar
ch
ical
Mu
lt
i
-
C
las
s
"
.
T
h
e
test
s
ca
r
r
ied
o
u
t
o
n
th
e
d
if
f
er
en
t
d
atab
ases
:
ND
-
co
n
tac
t,
C
asia
-
I
r
i
s
-
I
n
ter
v
al
a
n
d
C
a
s
ia
-
I
r
i
s
-
S
y
n
,
L
i
v
Det
-
I
r
is
-
2017
-
W
ar
s
a
w
s
h
o
w
ed
a
r
ec
o
g
n
itio
n
r
a
te
eq
u
al
to
1
0
0
% a
n
d
FA
R
=
0
%,
FR
R
=
0
%.
H.
G.
Da
w
a
y
et
al.
[
1
0
]
p
r
e
s
en
ted
a
n
e
w
m
e
th
o
d
f
o
r
d
etec
tin
g
t
h
e
p
u
p
il.
T
h
e
m
et
h
o
d
in
v
o
lv
e
s
s
ev
er
al
s
tep
s
,
t
h
e
m
o
s
t
i
m
p
o
r
t
an
t
o
f
w
h
ich
d
ep
en
d
s
o
n
t
h
e
d
if
f
er
en
ce
i
n
co
lo
r
an
d
in
ten
s
it
y
b
et
w
ee
n
t
h
e
p
u
p
il
an
d
its
n
ei
g
h
b
o
r
h
o
o
d
.
T
h
ese
ch
ar
ac
ter
is
tic
s
ar
e
v
er
y
i
m
p
o
r
ta
n
t
to
lo
ca
te
an
d
ex
tr
ac
t
th
e
p
u
p
il.
T
h
u
s
,
t
h
e
p
u
p
il
is
a
r
eg
io
n
o
f
v
er
y
h
ig
h
i
n
te
n
s
it
y
(
co
lo
r
)
co
m
p
ar
ed
to
its
n
eig
h
b
o
r
h
o
o
d
.
T
h
e
ex
p
er
i
m
en
t
al
r
esu
lt
s
s
h
o
w
ed
a
r
ec
o
g
n
itio
n
r
ate
eq
u
al
to
1
0
0
%.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
Du
a
l iri
s
a
u
th
en
tica
tio
n
s
ystem
u
s
in
g
d
ezer
t
s
ma
r
a
n
d
a
ch
e
t
h
eo
r
y
(
K
a
mel
Gh
a
n
em
Gh
a
lem
)
4705
I
n
p
u
r
p
o
s
e
to
i
m
p
r
o
v
e
a
n
o
v
er
all
p
er
f
o
r
m
a
n
ce
in
ter
m
s
o
f
r
ec
o
g
n
itio
n
r
ate
an
d
m
i
tig
ate
er
r
o
r
s
,
th
e
r
esear
ch
er
s
h
a
v
e
u
s
ed
m
o
r
e
th
an
o
n
e
b
io
m
etr
ic
tr
ait,
a
n
d
th
u
s
,
t
h
e
m
u
ltib
io
m
etr
ic
s
y
s
te
m
s
h
a
v
e
e
m
er
g
ed
.
Nu
m
er
o
u
s
m
u
lti
-
b
io
m
etr
ic
s
y
s
te
m
s
h
a
v
e
b
ee
n
d
ev
elo
p
ed
,
w
h
ich
f
u
s
io
n
is
m
ad
e
at
Ma
tch
in
g
s
co
r
e.
I
n
[
1
1
]
,
th
e
r
esear
c
h
er
s
p
r
o
p
o
s
ed
a
n
e
w
ap
p
r
o
ac
h
f
o
r
r
e
co
g
n
itio
n
u
s
i
n
g
b
o
th
ir
is
e
s
.
T
h
e
ir
is
is
s
eg
m
e
n
ted
u
s
i
n
g
th
e
C
a
n
n
y
f
ilter
an
d
Ho
u
g
h
tr
a
n
s
f
o
r
m
,
th
e
n
t
h
e
s
e
g
m
e
n
ted
ir
is
is
n
o
r
m
alize
d
b
y
J
.
Dau
g
m
an
’
s
r
u
b
b
er
s
h
ee
t
m
o
d
el.
T
h
e
ir
i
s
f
ea
t
u
r
e
ex
tr
ac
tio
n
is
c
ar
r
ied
o
u
t
u
s
i
n
g
co
n
v
o
lu
t
io
n
o
f
th
e
n
o
r
m
al
ized
ir
is
w
it
h
1
D
L
o
g
-
Gab
o
r
f
ilter
s
t
h
en
t
h
e
p
h
a
s
e
o
f
f
il
ter
ed
ir
is
is
q
u
a
n
tized
i
n
o
r
d
er
to
g
e
n
er
ate
a
b
in
ar
y
co
d
e.
T
h
e
Ha
m
m
i
n
g
d
is
tan
ce
is
u
s
ed
f
o
r
Ma
tc
h
in
g
.
T
h
e
Ma
tch
in
g
o
p
er
atio
n
co
n
s
is
t
s
i
n
co
m
p
ar
i
n
g
th
e
t
w
o
ir
i
s
f
ea
tu
r
e
v
ec
to
r
s
o
f
a
p
er
s
o
n
w
it
h
t
h
e
o
th
er
s
;
i
f
t
h
e
Ha
m
m
i
n
g
Dis
tan
ce
s
ar
e
les
s
th
a
n
t
h
e
t
h
r
esh
o
ld
th
en
t
h
e
p
er
s
o
n
i
s
id
en
tifie
d
.
T
h
e
ex
p
er
i
m
en
tal
r
es
u
lt
s
s
h
o
w
ed
a
g
o
o
d
r
ec
o
g
n
itio
n
r
ate
e
q
u
al
to
9
9
.
9
2
%
w
it
h
an
F
R
R
=
9
.
9
6
%,
w
h
ile
f
o
r
u
n
i
m
o
d
al
s
y
s
te
m
s
(
lef
t
ir
is
a
n
d
r
ig
h
t
ir
is
)
t
h
e
r
ec
o
g
n
it
io
n
r
ate
is
eq
u
a
l
to
9
9
.
8
7
%
w
it
h
a
n
FR
R
=
1
4
.
6
2
% a
n
d
FP
R
=
1
5
.
6
8
%.
I
n
[
1
2
]
,
th
e
au
th
o
r
s
p
r
esen
ted
th
e
f
r
a
m
e
w
o
r
k
f
o
r
m
u
lti
m
o
d
al
b
io
m
e
tr
ic
f
u
s
io
n
b
ased
o
n
th
e
u
n
ce
r
tai
n
t
y
co
n
ce
p
t
o
f
De
m
p
s
ter
-
S
h
af
er
t
h
eo
r
y
.
A
co
m
b
in
atio
n
o
f
q
u
alit
y
m
ea
s
u
r
es
an
d
th
e
ac
c
u
r
ac
y
o
f
class
i
f
ier
s
(
eq
u
al
er
r
o
r
r
ate)
ar
e
p
r
o
p
o
s
ed
to
en
co
d
e
th
e
u
n
ce
r
tain
t
y
co
n
ce
p
t
to
i
m
p
r
o
v
e
th
e
f
u
s
io
n
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
r
ev
ea
le
d
a
g
o
o
d
p
e
r
f
o
r
m
a
n
ce
w
it
h
an
E
E
R
eq
u
al
to
1
%.
R
.
D
w
i
v
ed
i
an
d
S.
De
y
[
1
3
]
p
r
o
p
o
s
ed
a
ca
n
ce
lab
le
m
u
ltib
i
o
m
e
tr
ic
s
y
s
te
m
u
s
i
n
g
s
co
r
e
le
v
el
f
u
s
io
n
.
T
h
e
f
u
s
io
n
o
f
s
co
r
es
w
a
s
ap
p
lied
b
y
MC
w
ei
g
h
t
in
g
at
f
i
r
s
t
lev
el
an
d
R
A
w
ei
g
h
ti
n
g
at
th
e
s
ec
o
n
d
lev
el.
T
h
e
co
m
p
ar
ativ
e
an
al
y
s
is
s
h
o
w
s
t
h
at
th
e
p
r
o
p
o
s
ed
f
u
s
io
n
m
et
h
o
d
o
u
tp
er
f
o
r
m
s
th
e
e
x
is
t
in
g
w
eig
h
ti
n
g
ap
p
r
o
ac
h
es.
I
n
th
is
w
o
r
k
s
,
w
e
e
x
tr
ac
t
o
n
l
y
th
e
i
n
ter
io
r
h
al
f
o
f
t
h
e
ir
is
d
is
c
r
ath
er
t
h
a
n
t
h
e
w
h
o
le
ir
is
d
is
c,
w
h
ic
h
co
n
tain
s
t
h
e
m
o
s
t
r
ele
v
an
t
i
n
f
o
r
m
at
io
n
a
n
d
it
is
le
s
s
a
f
f
ec
te
d
b
y
n
o
is
e.
I
n
ad
d
itio
n
,
w
e
co
m
b
in
ed
t
w
o
s
o
u
r
ce
s
o
f
i
n
f
o
r
m
atio
n
(
lef
t
ir
is
a
n
d
r
ig
h
t
ir
is
)
w
it
h
h
i
g
h
d
eg
r
ee
o
f
c
o
n
f
lic
t
u
s
in
g
Dez
er
t
S
m
ar
a
n
d
ac
h
e
T
h
eo
r
y
,
w
h
ic
h
s
o
lv
es
t
h
e
p
r
o
b
lem
o
f
h
ig
h
l
y
r
ed
is
tr
ib
u
ted
m
as
s
es
co
n
f
l
icts
ar
is
i
n
g
u
n
d
er
th
e
De
m
p
s
ter
s
h
a
f
er
th
eo
r
y
.
T
h
e
r
em
i
n
d
er
p
ap
er
is
o
r
g
an
ized
as
f
o
llo
w
s
:
th
e
r
esear
ch
m
e
th
o
d
is
d
escr
ib
ed
in
s
ec
t
io
n
2
,
R
esu
lt
s
an
d
An
al
y
s
i
s
ar
e
p
r
esen
ted
in
s
ec
t
i
o
n
3
,
co
n
clu
s
io
n
s
ar
e
p
r
o
v
id
ed
in
s
ec
tio
n
4
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
T
h
e
k
e
y
id
ea
o
f
o
u
r
w
o
r
k
i
s
to
ex
tr
ac
t
o
n
l
y
t
h
e
d
is
cr
i
m
i
n
an
t
in
f
o
r
m
a
tio
n
f
r
o
m
ir
i
s
te
x
tu
r
e
a
n
d
p
r
o
p
o
s
es
D
ez
er
t
S
m
ar
an
d
ac
h
e
T
h
eo
r
y
(
DS
m
T
)
at
s
co
r
e
lev
el
f
u
s
io
n
to
o
p
er
ate
u
n
d
er
u
n
ce
r
tain
t
y
i
n
g
o
al
to
ac
h
iev
e
a
g
o
o
d
p
er
f
o
r
m
a
n
ce
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
is
co
m
p
o
s
ed
o
f
f
o
u
r
m
ai
n
s
tag
e
s
:
p
r
ep
r
o
ce
s
s
in
g
,
f
ea
t
u
r
e
ex
tr
ac
tio
n
,
f
u
s
io
n
,
a
n
d
m
atc
h
i
n
g
.
2
.
1
.
P
re
pro
ce
s
s
ing
s
t
age
First,
t
h
e
ir
i
s
i
m
a
g
es
r
eq
u
ir
e
g
o
i
n
g
th
r
o
u
g
h
t
h
e
p
r
ep
r
o
ce
s
s
in
g
p
h
ase
in
cl
u
d
i
n
g
s
e
g
m
en
tatio
n
a
n
d
n
o
r
m
aliza
t
io
n
.
2
.
1
.
1
.
I
ris s
eg
m
ent
a
t
io
n
T
h
e
s
eg
m
e
n
tatio
n
o
f
ir
is
i
s
r
ea
lized
b
y
co
m
m
o
n
l
y
E
d
g
e
d
etec
to
r
m
et
h
o
d
: H
o
u
g
h
tr
a
n
s
f
o
r
m
.
-
H
o
u
g
h
tr
a
n
s
f
o
r
m
-
A
l
g
o
r
it
hm
-
G
en
er
ate
ed
g
e
m
ap
u
s
in
g
t
h
e
C
an
n
y
f
ilter
.
-
C
a
n
n
y
p
ar
a
m
eter
s
:
T
h
e
s
tan
d
ar
d
d
ev
iatio
n
o
f
Ga
u
s
s
ia
n
s
m
o
o
t
h
i
n
g
f
ilter
:
σ
=
2
.
W
eig
h
ti
n
g
f
o
r
v
er
tical
g
r
ad
ien
ts
=
0
.
W
eig
h
ti
n
g
f
o
r
h
o
r
izo
n
ta
l g
r
ad
i
en
ts
=1
.
-
I
n
cr
ea
s
e
co
n
tr
a
s
t in
d
ar
k
ir
i
s
r
eg
io
n
.
I
m
ag
e
g
a
m
m
a
v
al
u
e:
en
h
a
n
ce
th
e
co
n
tr
ast o
f
b
r
ig
h
t r
eg
io
n
s
:
γ
=
1
.
9
.
-
Dete
ct
p
ix
e
l c
o
r
r
esp
o
n
d
in
g
t
o
th
e
lo
ca
l
m
ax
i
m
a
Dis
ta
n
ce
in
p
i
x
e
l u
n
it
s
to
b
e
lo
o
k
ed
at
o
n
ea
ch
s
id
e
o
f
ea
ch
p
ix
el
w
h
en
d
eter
m
in
in
g
,
w
h
e
th
er
it is
a
lo
ca
l
m
ax
i
m
u
m
o
r
n
o
t:
d
=
1
.
5
.
-
B
in
ar
ize
ir
is
i
m
a
g
e
u
s
in
g
H
y
s
ter
esis
t
h
r
es
h
o
ld
in
g
.
L
o
w
th
r
e
s
h
o
ld
T
1
=
0
.
19
.
Hig
h
th
r
e
s
h
o
ld
T
2
=
0
.
20
.
T
h
e
Fig
u
r
e
1
illu
s
tr
ate
s
a
d
if
f
er
en
t
s
s
tep
o
f
Ho
u
g
h
tr
an
s
f
o
r
m
.
T
h
en
,
a
C
ir
cu
lar
Ho
u
g
h
T
r
an
s
f
o
r
m
d
etec
ts
at
f
ir
s
t
t
h
e
ir
is
/
s
cler
a
b
o
u
n
d
ar
y
an
d
th
e
ir
is
/p
u
p
il
b
o
u
n
d
ar
y
.
T
h
e
e
y
elas
h
es
ar
e
d
etec
ted
b
y
g
lo
b
al
th
r
es
h
o
ld
in
g
(
T
=
1
0
0
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
6
,
Dec
em
b
er
2
0
1
9
:
4
7
0
3
-
4
7
1
2
4706
Fig
u
r
e
1
.
Dif
f
er
en
t
s
s
tep
o
f
H
o
u
g
h
T
r
an
s
f
o
r
m
.
I
n
th
i
s
w
o
r
k
,
t
h
e
o
b
j
ec
tiv
e
is
t
o
ex
tr
ac
t
o
n
l
y
r
ele
v
a
n
t
i
n
f
o
r
m
atio
n
f
r
o
m
ir
is
,
w
h
ic
h
is
r
ep
r
esen
ted
b
y
th
e
s
tr
u
ct
u
r
al
v
ar
iatio
n
o
f
t
h
e
ir
is
te
x
t
u
r
e
(
h
i
g
h
g
r
ad
ie
n
t
ar
ea
s
)
,
o
n
l
y
t
h
e
in
ter
n
al
h
al
f
o
f
t
h
e
ir
is
d
is
c
i
s
ex
p
lo
ited
r
ath
er
th
an
w
h
o
le,
b
ec
au
s
e
it
co
n
ta
in
s
th
e
m
o
s
t
r
e
lev
an
t
in
f
o
r
m
at
io
n
[
1
4
]
an
d
it
is
less
a
f
f
ec
ted
b
y
th
e
n
o
is
e
as
s
h
o
w
n
in
Fi
g
u
r
e
2
.
I
n
d
ee
d
,
th
e
p
r
o
p
o
s
ed
tech
n
iq
u
e
d
ec
r
ea
s
es t
h
e
co
m
p
lex
it
y
a
n
d
th
e
co
m
p
u
ta
tio
n
l
o
ad
w
it
h
o
u
t
lo
s
i
n
g
in
f
o
r
m
a
tio
n
as
s
h
o
w
n
in
T
ab
le
1
.
Fro
m
T
ab
le
1
,
w
e
d
en
o
te
t
h
at
tr
ea
t
m
en
t
u
s
in
g
o
n
l
y
h
al
f
ir
is
d
is
c
i
s
m
o
r
e
e
f
f
icien
t
w
it
h
an
ac
c
u
r
ac
y
o
f
9
9
.
9
6
%
an
d
p
r
o
ce
s
s
in
g
ti
m
e
f
o
r
o
n
e
ir
is
i
m
ag
e
o
f
1
2
.
3
6
s
th
a
n
th
e
tr
ea
t
m
en
t
u
s
i
n
g
a
w
h
o
le
ir
i
s
d
is
c
w
it
h
ac
cu
r
ac
y
9
9
.
8
7
% o
f
an
d
p
r
o
ce
s
s
i
n
g
ti
m
e
o
f
2
2
.
8
8
s
.
r
hal
f
of
i
r
i
s
d
i
s
c
=
r
p
up
i
l
+
(
r
i
r
i
s
−
r
p
up
i
l
)
/
2
(
1
)
Fig
u
r
e
2
.
Deli
m
ita
tio
n
o
f
o
n
l
y
th
e
in
ter
n
al
h
al
f
o
f
ir
is
d
is
c
T
ab
le
1
.
C
o
m
p
ar
is
o
n
: T
o
tal
ir
is
d
is
c
v
s
Hal
f
ir
is
d
is
c
A
c
c
u
r
a
c
y
(
%)
P
r
o
c
e
ssi
n
g
t
i
me
f
o
r
o
n
e
i
r
i
s
i
mag
e
(
s)
T
o
t
a
l
i
r
i
s d
i
sc
9
9
.
8
7
2
2
.
8
8
H
a
l
f
i
r
i
s
d
i
s
c
9
9
.
9
6
1
2
.
3
6
2
.
1
.
2
.
I
ris no
r
m
a
l
iza
t
io
n
T
h
e
ir
is
d
is
c
d
o
es
n
o
t
a
l
w
a
y
s
h
a
v
e
t
h
e
s
a
m
e
d
i
m
en
s
io
n
,
e
v
en
f
o
r
e
y
e
i
m
ag
e
s
o
f
th
e
s
a
m
e
p
er
s
o
n
;
th
is
i
s
d
u
e
to
v
ar
io
u
s
p
r
o
b
lem
s
as f
o
llo
w
s
:
a)
Dif
f
er
en
t a
cq
u
is
it
io
n
co
n
d
itio
n
s
o
f
t
h
e
e
y
e
i
m
a
g
es.
Di
latio
n
an
d
co
n
tr
ac
tio
n
o
f
t
h
e
p
u
p
il d
u
e
to
th
e
v
ar
iatio
n
o
f
t
h
e
ill
u
m
i
n
atio
n
le
v
el.
b)
T
h
e
cir
cles o
f
ir
is
an
d
p
u
p
il a
r
e
n
o
t c
o
n
ce
n
tr
ic.
I
n
o
r
d
er
to
o
v
er
co
m
e
th
e
s
e
p
r
o
b
lem
s
,
a
s
ta
g
e
o
f
n
o
r
m
aliza
ti
o
n
is
ap
p
lied
.
I
t
co
n
s
is
t
s
o
f
tr
a
n
s
f
o
r
m
i
n
g
th
e
r
eg
io
n
o
f
th
e
ir
is
d
is
c
to
r
ec
tify
t
h
e
d
i
m
e
n
s
i
o
n
s
o
f
all
th
e
ir
is
d
is
cs,
b
y
u
s
i
n
g
th
e
h
o
m
o
g
en
o
u
s
r
u
b
b
er
s
h
e
e
t
m
o
d
el
p
r
o
p
o
s
ed
b
y
Dau
g
m
a
n
[
1
]
.
I
t
tr
an
s
f
o
r
m
s
ea
ch
p
o
in
t
in
th
e
ir
is
ar
ea
to
th
e
p
o
lar
co
o
r
d
in
ates
(
r
,
θ
)
,
w
h
er
e
r
is
o
n
t
h
e
in
ter
v
al
[
0
,
1
]
an
d
θ
is
an
g
le
[
0
,
2
π
]
,
as illu
s
tr
ated
i
n
Fi
g
u
r
e
3
.
I
n
o
u
r
s
y
s
te
m
,
(
2
0
*
2
4
0
)
p
o
in
ts
w
er
e
u
s
ed
,
b
u
t
o
n
l
y
(
1
0
*
2
4
0
)
p
o
in
ts
co
r
r
esp
o
n
d
in
g
to
t
h
e
i
n
ter
n
a
l
h
al
f
o
f
t
h
e
ir
i
s
d
is
c
t
h
at
co
n
tai
n
s
th
e
m
o
s
t
r
ele
v
a
n
t
i
n
f
o
r
m
at
io
n
a
n
d
w
h
ic
h
i
s
less
af
f
ec
ted
b
y
n
o
is
e,
ar
e
r
etain
ed
f
o
r
th
e
n
e
x
t step
s
o
f
th
e
p
r
o
ce
s
s
i
n
g
,
as
s
h
o
w
n
i
n
Fig
u
r
e
4
.
I
r
i
s i
ma
g
e
Ed
g
e
ma
p
I
n
c
r
e
a
se
c
o
n
t
r
a
st
N
o
n
-
max
i
ma
su
p
p
r
e
ssi
o
n
B
i
n
a
r
y
i
mag
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
Du
a
l iri
s
a
u
th
en
tica
tio
n
s
ystem
u
s
in
g
d
ezer
t
s
ma
r
a
n
d
a
ch
e
t
h
eo
r
y
(
K
a
mel
Gh
a
n
em
Gh
a
lem
)
4707
Fig
u
r
e
3
.
Dau
g
m
a
n
r
u
b
b
er
s
h
e
et
m
o
d
el
[
1
]
Fig
u
r
e
4
.
No
r
m
aliza
tio
n
o
f
t
h
e
s
eg
m
en
ted
ir
is
2
.
2
.
F
ea
t
ure
ex
t
ra
ct
io
n sta
g
e
Af
ter
t
h
at,
t
h
e
f
ea
t
u
r
e
e
x
tr
ac
tio
n
s
tag
e
is
ap
p
lied
i
n
p
u
r
p
o
s
e
to
ex
tr
ac
t
th
e
m
o
s
t
d
is
cr
i
m
i
n
atio
n
in
f
o
r
m
atio
n
p
r
ese
n
t i
n
th
e
ir
i
s
r
eg
io
n
.
Fo
r
th
i
s
r
ea
s
o
n
,
a
b
en
c
h
o
f
t
w
o
1
D
L
o
g
-
Gab
o
r
f
ilter
is
u
s
ed
.
a)
1
D
lo
g
-
g
a
bo
r
f
ilte
r
-
T
h
e
Fas
t Fo
u
r
ier
T
r
an
s
f
o
r
m
i
s
ap
p
lied
f
o
r
ea
ch
lin
e
o
f
th
e
n
o
r
m
alize
d
m
atr
i
x
i
m
ag
e
(
F
FT
to
1
D
s
ig
n
al
s
)
.
-
T
h
en
,
th
e
I
n
v
er
s
e
Fa
s
t
Fo
u
r
ier
T
r
an
s
f
o
r
m
I
FF
T
is
ap
p
lied
o
n
th
e
m
u
lt
ip
licatio
n
F
FT
(
1
D
s
ig
n
al
s
)
b
y
a
1
D
L
o
g
-
Gab
o
r
Fil
ter
.
-
T
h
e
f
r
eq
u
en
c
y
r
e
s
p
o
n
s
e
o
f
a
1
D
L
o
g
-
Gab
o
r
f
ilter
i
s
g
i
v
en
b
y
:
(
)
=
e
xp
(
−
l
og
(
0
)
2
l
og
(
0
)
2
)
(
2
)
b)
P
a
ra
m
et
er
s
set
ting
-
A
b
en
c
h
o
f
t
w
o
1
D
L
o
g
-
Gab
o
r
f
ilter
s
is
u
s
ed
.
-
T
h
e
s
tan
d
ar
d
d
ev
iatio
n
o
f
t
h
e
1
D
L
o
g
-
Gab
o
r
w
a
v
elet
is
g
i
v
e
n
b
y
=
2
.
-
T
h
e
ce
n
ter
f
r
eq
u
en
c
y
o
f
t
h
e
1
D
L
o
g
-
Gab
o
r
w
a
v
elet
i
s
g
i
v
e
n
b
y
0
=
0
.
05
I
n
d
ee
d
,
th
e
p
h
ase
o
f
a
f
ilter
ed
im
a
g
e
w
as
q
u
an
tized
u
s
i
n
g
f
o
u
r
-
q
u
ad
r
a
n
ts
o
f
Dau
g
m
a
n
[
1
]
,
w
h
e
n
g
o
in
g
f
r
o
m
o
n
e
q
u
ad
r
an
t
to
a
n
ad
j
ac
en
t
q
u
ad
r
an
t,
o
n
e
b
it
i
s
ch
a
n
g
ed
as
s
h
o
w
n
i
n
Fi
g
u
r
e
5
.
T
h
e
en
co
d
in
g
p
r
o
ce
s
s
p
r
o
d
u
ce
s
a
b
it
w
i
s
e
te
m
p
late
co
n
tai
n
i
n
g
a
n
u
m
b
er
o
f
in
f
o
r
m
at
io
n
b
its
as
s
h
o
w
n
in
Fig
u
r
e
6
,
th
e
to
tal
n
u
m
b
er
o
f
b
it
s
i
n
t
h
e
te
m
p
late
(
9
6
0
0
b
its
)
w
ill b
e
t
h
e
a
n
g
u
la
r
r
eso
lu
tio
n
(
2
4
0
)
tim
e
s
t
h
e
r
a
d
ial
r
eso
lu
tio
n
(
1
0
)
,
ti
m
e
s
2
,
ti
m
es t
h
e
n
u
m
b
er
o
f
f
ilter
s
u
s
ed
(
2
)
.
Fig
u
r
e
5
.
Qu
an
t
izatio
n
p
h
a
s
e
[
1
]
Fig
u
r
e
6
.
I
r
is
co
d
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
6
,
Dec
em
b
er
2
0
1
9
:
4
7
0
3
-
4
7
1
2
4708
2
.
3
.
M
a
t
ching
s
t
a
g
e
[
Co
mm
ent
:
M
a
t
ching
s
co
re
co
m
e
b
ef
o
re
F
us
io
n sta
g
e
]
T
h
e
m
a
tch
i
n
g
co
n
s
is
t
s
in
co
m
p
ar
in
g
t
w
o
ir
is
co
d
e
u
s
i
n
g
Ha
m
m
in
g
d
is
tan
ce
.
T
h
e
Ha
m
m
i
n
g
Dis
tan
ce
(
HD)
is
d
ef
in
ed
b
y
:
HD
=
∑
X
j
⨁
Y
j
N
j
=
1
(
3
)
W
h
er
e
X
j
an
d
Y
j
ar
e
th
e
t
w
o
b
it
w
is
e
ir
i
s
co
d
e,
N
is
t
h
e
n
u
m
b
er
o
f
b
it
s
i
n
ea
ch
ir
is
co
d
e.
L
iter
all
y
,
th
e
Ha
m
m
i
n
g
d
is
tan
ce
ca
lcu
l
ates
th
e
n
u
m
b
er
o
f
d
if
f
er
e
n
t
an
d
v
alid
b
its
f
o
r
th
e
t
w
o
ir
is
co
d
e
b
et
w
ee
n
X
j
an
d
Y
j
.
T
h
e
n
u
m
b
er
o
f
tr
a
n
s
latio
n
b
its
t
h
at
co
m
p
e
n
s
ate
s
t
h
e
r
o
tatio
n
o
f
t
h
e
ir
is
n
ee
d
s
to
b
e
f
i
x
ed
.
W
e
ap
p
lied
a
tr
an
s
latio
n
o
f
t
h
e
ir
is
co
d
e
in
an
in
ter
v
al
[
−
,
+
]
b
its
.
W
e
tak
e
in
to
co
n
s
id
er
atio
n
t
h
e
m
i
n
i
m
u
m
Ha
m
m
i
n
g
d
is
tan
ce
.
2
.
4
.
F
us
io
n sta
g
e
I
n
t
h
is
s
ta
g
e,
s
co
r
e
le
v
el
f
u
s
i
o
n
u
s
i
n
g
Dez
er
t
S
m
ar
a
n
d
ac
h
e
th
eo
r
y
(
DS
m
T
)
w
as
ap
p
lied
o
n
g
o
al
to
i
m
p
r
o
v
e
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
d
u
al
ir
is
s
y
s
te
m
.
S
co
r
e
leve
l fu
s
io
n
Ma
tch
i
n
g
s
co
r
e
lev
el
f
u
s
io
n
c
o
m
b
i
n
es
t
h
e
s
co
r
es
g
e
n
er
ated
b
y
m
u
ltip
le
class
i
f
ier
s
r
elatin
g
to
th
e
lef
t
an
d
r
ig
h
t
ir
i
s
t
o
af
f
ir
m
t
h
e
v
e
r
ac
it
y
o
f
th
e
clai
m
ed
id
en
tit
y
.
T
h
e
Dez
er
t
Sm
ar
an
d
ac
h
e
th
e
o
r
y
o
p
er
ates
u
n
d
er
h
y
p
er
p
o
w
er
s
et
D
Θ
.
T
h
u
s
,
DSm
T
is
ab
le
to
f
u
n
ctio
n
p
r
o
p
er
l
y
n
o
t
o
n
l
y
w
it
h
th
e
u
n
io
n
s
b
u
t
also
w
ith
in
ter
s
ec
tio
n
s
.
D
s
m
C
las
s
ic
h
a
s
co
m
b
i
n
atio
n
r
u
le
[
1
5
-
1
7
]
:
m
(
C
)
=
∑
m
1
(
A
)
m
2
(
B
)
;
A
∩
B
=
C
A
,
B
∈
D
Θ
,
∀
C
∈
D
Θ
(
4
)
E
x
a
m
p
le
Θ
=
{
S
l
ef
t
,
S
r
i
g
ht
}
(
5
)
D
Θ
=
{
∅
,
S
l
ef
t
,
S
r
i
g
h
t
,
S
l
ef
t
∪
S
r
i
g
ht
,
S
l
ef
t
∩
S
r
i
g
ht
}
(
6
)
∅
:
E
m
p
t
y
s
et
S
l
ef
t
:
H
y
p
o
th
es
is
as
s
u
m
i
n
g
th
a
t t
wo
in
d
iv
id
u
al
s
h
a
v
e
s
a
m
e
le
f
t ir
is
.
S
Ri
g
ht
:
H
y
p
o
th
es
is
as
s
u
m
i
n
g
th
a
t t
wo
in
d
iv
id
u
al
s
h
a
v
e
a
s
a
m
e
r
ig
h
t
ir
is
.
S
l
ef
t
∪
S
r
i
g
h
t
∶
H
y
p
o
th
esi
s
as
s
u
m
i
n
g
th
at
t
w
o
in
d
iv
id
u
als
h
av
e
d
i
f
f
er
en
t ir
is
S
l
ef
t
∩
S
r
i
g
h
t
∶
H
y
p
o
th
es
is
as
s
u
m
i
n
g
th
a
t t
w
o
in
d
iv
id
u
als
h
a
v
e
d
if
f
er
en
t ir
i
s
.
2
.
5
.
Dec
is
io
n
T
h
e
d
ec
is
io
n
is
m
ad
e
b
y
f
i
x
in
g
a
t
h
r
esh
o
ld
.
T
h
e
t
w
o
ir
is
es
c
o
m
p
ar
ed
w
ill b
e
co
n
s
id
er
ed
as
b
elo
n
g
i
n
g
to
th
e
s
a
m
e
p
er
s
o
n
if
t
h
e
ca
lc
u
lated
s
co
r
e
is
in
f
er
io
r
to
a
th
r
e
s
h
o
ld
.
3.
RE
SU
L
T
S
AND
AN
AL
Y
SI
S
3
.
1
.
Si
m
ula
t
io
n e
nv
iro
n
m
en
t
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
h
as
b
ee
n
test
ed
o
n
a
s
u
b
s
et
o
f
ir
is
d
a
tab
ase
C
ASI
A
-
I
r
is
V3
-
I
n
ter
v
al
[
1
8
]
in
o
r
d
er
to
ev
alu
ate
it
s
p
er
f
o
r
m
an
ce
i
n
au
th
e
n
ticat
io
n
m
o
d
e.
T
h
e
s
u
b
s
et
co
n
tai
n
s
1
1
8
0
ey
e
i
m
ag
e
s
o
f
1
1
8
in
d
iv
id
u
als
(
cla
s
s
es),
a
n
d
ea
ch
in
d
i
v
id
u
al
h
as
f
i
v
e
ir
is
s
a
m
p
les
f
o
r
th
e
le
f
t
e
y
e
a
n
d
f
i
v
e
ir
is
s
a
m
p
les
f
o
r
th
e
r
ig
h
t e
y
e.
3
.
2
.
P
er
f
o
rm
a
nce
m
et
rics
-
Fals
e
R
ej
ec
t Rate
(
F
R
R
)
: a
ls
o
k
n
o
w
n
as T
y
p
e
I
er
r
o
r
,
is
th
e
m
ea
s
u
r
e
o
f
t
h
e
p
r
o
b
ab
ilit
y
t
h
at
th
e
b
io
m
etr
ic
s
ec
u
r
it
y
s
y
s
te
m
w
il
l in
co
r
r
ec
tl
y
r
ej
ec
t a
n
ac
ce
s
s
atte
m
p
t b
y
a
n
au
t
h
o
r
ized
u
s
er
.
-
Fals
e
A
cc
ep
t
R
ate
(
F
A
R
)
:
also
k
n
o
w
n
as
T
y
p
e
I
I
er
r
o
r
,
is
t
h
e
m
ea
s
u
r
e
o
f
t
h
e
p
r
o
b
ab
ilit
y
th
a
t
th
e
b
io
m
etr
ic
s
ec
u
r
it
y
s
y
s
te
m
w
il
l in
co
r
r
ec
tl
y
ac
ce
p
t a
n
ac
ce
s
s
atte
m
p
t b
y
an
u
n
a
u
t
h
o
r
ized
u
s
er
.
-
E
E
R
(
E
q
u
al
E
r
r
o
r
R
ate)
:
T
h
e
E
E
R
is
th
e
o
p
er
atin
g
p
o
in
t
f
o
r
w
h
ic
h
t
h
e
Fals
e
R
ej
ec
t
R
a
te
(
F
R
R
)
is
eq
u
al
to
th
e
Fals
e
A
cc
ep
t
R
a
te
(
FP
R
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
Du
a
l iri
s
a
u
th
en
tica
tio
n
s
ystem
u
s
in
g
d
ezer
t
s
ma
r
a
n
d
a
ch
e
t
h
eo
r
y
(
K
a
mel
Gh
a
n
em
Gh
a
lem
)
4709
3
.
3
.
Dec
ida
bil
it
y
Dec
id
ab
ilit
y
[
1
]
i
s
t
h
e
b
e
s
t
m
etr
ic
w
h
ich
i
n
d
ee
d
tak
e
s
i
n
to
ac
co
u
n
t
th
e
m
ea
n
a
n
d
s
ta
n
d
ar
d
d
ev
iatio
n
o
f
th
e
i
n
tr
a
-
c
lass
a
n
d
in
ter
-
cla
s
s
d
is
tr
ib
u
tio
n
s
:
d
′
=
|
μ
s
−
μ
d
|
√
(
σ
s
2
+
σ
d
2
)
2
(
7
)
Dec
id
ab
ilit
y
d
'
i
s
a
d
is
ta
n
ce
in
s
tan
d
ar
d
d
ev
iat
io
n
s
ca
lc
u
l
ated
u
s
i
n
g
(
7
)
,
w
h
ic
h
i
s
a
f
u
n
ct
io
n
o
f
th
e
m
a
g
n
it
u
d
e
o
f
th
e
d
if
f
er
en
c
e
b
et
w
ee
n
t
h
e
m
ea
n
o
f
th
e
i
n
tr
a
-
clas
s
d
is
tr
ib
u
tio
n
μ
s
,
an
d
th
e
m
ea
n
o
f
t
h
e
in
ter
-
class
d
is
tr
ib
u
tio
n
μ
d
,
th
e
s
tan
d
ar
d
d
ev
iatio
n
o
f
th
e
i
n
tr
a
-
cl
ass
an
d
i
n
ter
-
cla
s
s
d
is
tr
ib
u
ti
o
n
s
,
σ
s
an
d
σ
d
r
esp
ec
tiv
el
y
.
T
ab
le
2
.
Dis
cid
a
b
ilit
y
tab
le
f
o
r
v
ar
io
u
s
n
u
m
b
er
s
o
f
b
it
-
s
h
i
f
ts
N
u
mb
e
r
o
f
sh
i
f
t
s
μ
s
σ
s
μ
d
σ
d
d
′
0
0
.
3
3
0
0
0
.
0
7
2
3
0
.
4
9
1
4
0
.
0
2
8
4
3
.
4
3
1
4
1
0
.
3
1
3
7
0
.
0
6
9
7
0
.
4
8
6
0
0
.
0
2
7
9
3
.
8
1
4
9
2
0
.
3
0
7
2
0
.
0
6
6
8
0
.
4
8
1
2
0
.
0
2
6
9
4
.
0
2
6
4
3
0
.
3
0
4
4
0.
0
6
5
3
0
.
4
7
7
2
0
.
0
2
5
8
4
.
0
7
4
2
4
0
.
3
0
3
2
0
.
0
6
4
6
0
.
4
7
3
8
0
.
0
2
4
7
4
.
0
4
3
1
5
0
.
3
0
2
8
0
.
0
6
4
2
0
.
4
7
0
9
0
.
0
2
3
8
3
.
9
9
0
7
6
0
.
3
0
2
5
0
.
0
6
3
9
0
.
4
6
8
4
0
.
0
2
3
0
3
.
9
3
6
2
7
0
.
0
6
3
7
0
.
0
6
4
2
0
.
0
2
2
3
0
.
0
2
1
6
3
.
8
8
6
2
8
0
.
3
0
2
3
0
.
0
6
3
5
0
.
4
6
4
5
0
.
0
2
1
6
3
.
8
3
0
3
9
0
.
3
0
2
2
0
.
0
6
3
4
0
.
4
6
2
9
0
.
0
2
1
1
3
.
7
9
9
9
1
0
0
.
2
7
5
8
0
.
0
6
3
9
0
.
4
6
4
3
0
.
0
2
0
1
4
.
3
9
6
0
Fig
u
r
e
7
.
Dec
id
ab
ilit
y
cu
r
v
e
f
o
r
v
ar
io
u
s
n
u
m
b
er
s
o
f
b
it
-
s
h
if
t
s
Usi
n
g
(
7
)
,
s
ev
er
al
d
if
f
er
en
t
d
ec
id
ab
ilit
y
ar
e
f
o
u
n
d
o
u
t
u
s
in
g
0
-
b
it
s
h
if
t
to
1
0
-
b
it
s
h
i
f
t
to
w
ar
d
s
b
o
t
h
lef
t
an
d
r
ig
h
t
ir
is
te
m
p
lates
.
T
h
e
h
i
g
h
er
d
ec
id
ab
ilit
y
i
s
eq
u
al
to
4
.
0
7
4
2
at
3
b
it
s
h
if
t
as
s
h
o
w
n
in
T
ab
le
2
an
d
Fig
u
r
e
7
t
h
at
g
u
ar
a
n
tees
g
o
o
d
s
ep
ar
atio
n
o
f
in
tr
a
-
cla
s
s
a
n
d
in
ter
-
c
lass
d
is
tr
ib
u
tio
n
s
,
w
h
ich
allo
w
s
f
o
r
m
o
r
e
ac
cu
r
ate
r
ec
o
g
n
it
io
n
3
.
4
.
Sco
re
le
v
el
f
us
io
n
I
n
f
ac
t,
w
e
ca
lc
u
lated
t
h
e
f
u
s
i
o
n
s
co
r
e
u
s
i
n
g
Ha
m
i
n
g
d
i
s
ta
n
ce
s
o
b
tai
n
ed
b
y
co
m
p
ar
i
n
g
t
h
e
i
n
d
iv
id
u
als
f
r
o
m
t
h
eir
ir
is
HD
L
:
Ha
m
m
i
n
g
d
is
tan
ce
o
b
tain
ed
b
y
co
m
p
ar
in
g
t
h
e
in
d
i
v
id
u
al
s
f
r
o
m
t
h
eir
le
f
t ir
is
.
S
L
:
Sco
r
e
o
b
tain
ed
b
y
co
m
p
ar
in
g
t
h
e
in
d
i
v
id
u
al
s
f
r
o
m
th
e
ir
lef
t i
r
is
.
HD
R
:
Ha
m
m
i
n
g
d
is
tan
ce
o
b
tain
ed
b
y
co
m
p
ar
in
g
t
h
e
in
d
i
v
id
u
al
s
f
r
o
m
t
h
eir
r
ig
h
t ir
i
s
.
S
R
:
Sco
r
e
o
b
tain
ed
b
y
co
m
p
ar
in
g
t
h
e
in
d
i
v
id
u
al
s
f
r
o
m
th
e
ir
r
ig
h
t
ir
is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
6
,
Dec
em
b
er
2
0
1
9
:
4
7
0
3
-
4
7
1
2
4710
Alg
o
rit
h
m
F
o
r
e
a
c
h
i
n
d
iv
id
u
a
l
in
d
v
F
o
r
e
a
c
h
d
if
fe
re
n
t
iri
s
,
:
su
c
h
a
s
,
b
e
lo
n
g
to
iri
s se
t
o
f
in
d
iv
id
u
a
l
in
d
v
Ca
lcu
late
th
e
sc
o
re
S
L
(
i
,
j
)
=
1
−
HD
L
(
i
,
j
)
Ca
lcu
late
th
e
sc
o
re
S
R
(
i
,
j
)
=
1
−
HD
R
(
i
,
j
)
Ca
lcu
late
th
e
f
u
sio
n
o
f
sc
o
re
S
f
(
i
,
j
)
=
S
L
(
i
,
j
)
×
S
R
(
i
,
j
)
k
=
1
F
o
r
s
=
0
:
0
.
05
:
1
If
S
f
(
i
,
j
)
<
th
e
n
FN
(
k
)
=
FN
(
k
)
+
1
%
F
a
lse
Ne
g
a
ti
f
k
=
k
+
1
E
n
d
if
En
d
En
d
En
d
F
o
r
e
a
c
h
d
if
fe
re
n
t
in
d
iv
id
u
a
l
,
F
o
r
e
a
c
h
d
if
f
e
r
e
n
t
iri
s
(
,
)
su
c
h
a
s
i
b
e
lo
n
g
to
iri
s se
t
o
f
in
d
iv
id
u
a
ls
i
n
d
v
i
a
n
d
j
b
e
l
o
n
g
to
iri
s se
t
o
f
in
d
i
v
id
u
a
ls
in
d
v
j
Ca
lcu
late
th
e
sc
o
re
S
L
(
i
,
j
)
=
1
−
HD
L
(
i
,
j
)
Ca
lcu
late
th
e
sc
o
re
S
R
(
i
,
j
)
=
1
−
HD
R
(
i
,
j
)
Ca
lcu
late
th
e
f
u
sio
n
o
f
sc
o
re
S
f
(
i
,
j
)
=
S
L
(
i
,
j
)
×
S
R
(
i
,
j
)
k
=
1
F
o
r
s
=
0
:
0
.
05
:
1
If
S
f
(
i
,
j
)
≥
s
th
e
n
F
P
(
k
)
=
FP
(
k
)
+
1
%
F
a
lse
P
o
sit
if
k
=
k
+
1
En
d
if
E
n
d
En
d
En
d
m
a
x
in
d
v
=
n
u
m
b
e
r
o
f
in
d
iv
id
u
a
ls
n
b
tr=n
u
m
b
e
r
o
f
iri
s im
a
g
e
s p
e
r
in
d
iv
id
u
a
l
nbi
nter
=
m
a
x
in
d
v
×
(
nbt
r
×
(
nbt
r
−
1
)
∕
2
)
nbi
ntr
a
=
m
a
x
in
d
v
×
(
nbt
r
×
(
nbt
r
−
1
)
/
2
)
TN
=
nbi
nter
−
FP
%
T
ru
e
Ne
g
a
ti
f
TP
=
nbi
ntr
a
−
FN
%
T
ru
e
P
o
sit
if
T
P
R
=
(
TP
∕
nbi
ntr
a
)
×
100
%
T
ru
e
P
o
siti
f
Ra
te
T
NR
=
(
TN
∕
nbi
nter
)
×
100
%
T
ru
e
Ne
g
a
ti
f
R
a
te
F
A
R
=
(
FP
/
nbi
nter
)
×
100
%
F
a
lse
A
c
c
e
p
t
Ra
te
F
R
R
=
(
FN
/
nbi
ntr
a
)
×
100
%
F
a
lse
Re
jec
t
Ra
te
A
ccu
ra
cy
=
(
(
TP
+
TN
)
/
(
nbi
ntr
a
+
nbi
nter
)
)
×
100
T
h
e
d
u
al
ir
is
s
y
s
te
m
u
s
in
g
D
s
m
C
at
s
co
r
e
le
v
el
f
u
s
io
n
r
ea
c
h
es
an
ac
c
u
r
ac
y
r
ate
o
f
9
9
.
9
6
%
an
d
F
AR
o
f
0
%,
FR
R
o
f
3
.
8
9
%,
EER
o
f
2
%
as
s
h
o
w
n
i
n
Fi
g
u
r
e
8
an
d
Fig
u
r
e
9
.
W
e
c
o
n
clu
d
e
f
r
o
m
Fi
g
u
r
e
1
0
th
at
th
e
R
OC
(
R
ec
eiv
er
Op
er
ati
n
g
C
h
ar
ac
ter
is
t
ic)
o
f
d
u
al
ir
is
s
y
s
te
m
u
s
i
n
g
Ds
m
C
a
t
s
co
r
e
lev
el
f
u
s
io
n
f
it
th
e
o
r
ig
i
n
,
w
h
ic
h
p
r
o
v
es t
h
e
p
er
f
o
r
m
an
ce
o
f
o
u
r
m
et
h
o
d
.
3
.
5
.
Co
m
pa
ri
s
o
n o
f
v
a
rio
us
a
pp
ro
a
ches
W
e
d
en
o
te
f
r
o
m
T
ab
le
3
,
th
at
th
e
p
r
o
p
o
s
ed
d
u
al
ir
is
a
u
t
h
en
t
icatio
n
s
y
s
te
m
g
iv
e
s
a
c
o
m
p
eti
tiv
e
p
er
f
o
r
m
a
n
ce
w
i
th
a
cc
u
r
ac
y
o
f
9
9
.
9
6
%
,
FAR
o
f
0
%
,
F
R
R
o
f
3
.
8
9
%
,
E
E
R
o
f
2
%
i
n
co
m
p
ar
i
s
o
n
w
i
th
o
th
er
ap
p
r
o
ch
s
.
T
h
e
p
r
o
b
lem
o
f
De
m
p
s
ter
Sh
a
f
er
t
h
eo
r
y
u
s
ed
in
DST
ap
p
r
o
ac
h
[
1
2
]
(
E
E
R
o
f
1
%)
th
at
co
n
s
is
t
s
o
f
co
m
b
i
n
i
n
g
t
w
o
s
o
u
r
ce
s
o
f
i
n
f
o
r
m
atio
n
w
i
th
h
i
g
h
d
eg
r
ee
o
f
c
o
n
f
lic
t
is
r
eso
l
v
ed
.
T
h
e
p
r
o
p
o
s
ed
m
e
th
o
d
is
b
ased
o
n
Dez
er
t
S
m
ar
a
n
d
ac
h
e
C
la
s
s
ic
r
u
le
(
Ds
m
C
)
th
at
s
o
l
v
es
th
is
p
r
o
b
le
m
.
I
f
tak
h
ar
a
n
d
al
[
1
1
]
u
s
ed
t
h
e
f
u
s
io
n
m
et
h
o
d
b
ased
o
n
th
e
A
ND
r
u
le
th
at
g
iv
e
s
an
ac
cu
ar
ac
y
o
f
9
9
.
9
2
%
,
FA
R
o
f
0
%,
FR
R
o
f
9
.
9
6
%,
w
h
ich
i
s
m
o
r
e
d
r
asti
c
a
n
d
lead
s
to
i
m
p
r
o
v
e
th
e
F
AR
.
R
.
D
w
i
n
d
i
a
n
d
S.De
y
o
b
tai
n
ed
less
p
er
f
o
r
m
an
ce
i
n
ter
m
o
f
ac
cu
r
ac
y
eq
u
al
to
9
8
.
8
9
%,
E
E
R
o
f
0
.
6
9
%,
w
h
ich
u
s
ed
a
s
co
r
e
f
u
s
io
n
m
e
th
o
d
s
li
k
es:
M
C
w
eig
h
ti
n
g
an
d
R
A
w
ei
g
h
ti
n
g
.
T
h
ese
t
w
o
m
eth
o
d
s
u
s
in
g
f
o
r
o
p
ti
m
izatio
n
g
i
v
es
a
litt
le
i
m
p
r
o
v
e
m
e
n
t
to
th
e
p
er
f
o
r
m
a
n
ce
o
f
s
y
s
te
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
Du
a
l iri
s
a
u
th
en
tica
tio
n
s
ystem
u
s
in
g
d
ezer
t
s
ma
r
a
n
d
a
ch
e
t
h
eo
r
y
(
K
a
mel
Gh
a
n
em
Gh
a
lem
)
4711
Fig
u
r
e
8
.
T
h
e
ac
cu
r
ac
y
o
f
t
h
e
d
u
al
ir
is
s
y
s
te
m
Fig
u
r
e
9
.
FR
R
a
n
d
F
A
R
o
f
th
e
d
u
al
ir
is
s
y
s
te
m
Fig
u
r
e
1
0
.
R
OC
o
f
t
h
e
d
u
al
ir
i
s
s
y
s
te
m
T
ab
le
3
.
A
cc
u
r
ac
y
a
n
d
er
r
o
r
f
o
r
v
aio
u
s
ap
p
r
o
ac
h
es
I
r
i
s re
c
o
g
n
i
t
i
o
n
sy
st
e
ms
A
c
c
u
r
a
c
y
(
%)
F
A
R
(
%)
F
R
R
(
%)
EER
(
%)
I
f
t
a
k
h
a
r
a
n
d
A
sh
r
a
f
u
l
A
p
p
r
o
a
c
h
[
1
1
]
9
9
.
9
2
0
.
0
0
9
.
9
6
-
D
S
T
a
p
p
r
o
a
c
h
[
1
2
]
-
-
-
1
R
.
D
w
i
v
e
d
i
a
n
d
S
.
D
e
y
[
1
3
]
9
8
.
8
9
-
-
0
.
6
9
P
r
o
p
o
sal
d
u
a
l
i
r
i
s
a
u
t
h
e
n
t
i
c
a
t
i
o
n
sy
st
e
m
9
9
.
9
6
0
.
0
0
3
.
8
9
2
4.
CO
NCLU
SI
O
N
T
h
e
p
u
r
p
o
s
e
o
f
th
i
s
w
o
r
k
w
a
s
to
f
in
d
o
u
t
a
d
u
a
l
ir
is
a
u
t
h
e
n
ticatio
n
s
y
s
te
m
th
a
t
g
u
ar
an
te
es
a
g
o
o
d
p
er
f
o
r
m
a
n
ce
a
n
d
to
m
a
k
e
s
u
r
e
t
h
at
t
h
er
e
i
s
n
o
f
a
ls
e
a
cc
ep
tan
c
e
r
ate,
w
h
ic
h
p
r
o
m
i
s
es
u
s
e
f
u
l
s
ec
u
r
it
y
ap
p
licatio
n
s
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
co
n
s
is
t
s
in
s
e
g
m
en
tin
g
,
to
n
o
r
m
alizi
n
g
,
ch
ar
ac
ter
izi
n
g
an
d
en
co
d
in
g
th
e
ir
is
.
Fo
r
th
e
s
e
g
m
en
ta
tio
n
p
ar
t,
th
e
d
etec
tio
n
o
f
t
h
e
ir
is
/
p
u
p
il
cir
cles
w
as
p
er
f
o
r
m
ed
b
y
Ho
u
g
h
cir
c
u
lar
tr
an
s
f
o
r
m
.
On
l
y
t
h
e
in
ter
io
r
h
alf
o
f
t
h
e
ir
is
d
is
c
co
n
tai
n
i
n
g
t
h
e
m
o
s
t
r
elev
a
n
t
in
f
o
r
m
at
io
n
an
d
less
a
f
f
ec
ted
b
y
n
o
is
e,
w
h
ic
h
r
ed
u
ce
s
ti
m
e
co
m
p
lex
i
t
y
w
as
e
x
tr
ac
ted
.
I
r
is
n
o
r
m
al
izatio
n
p
ar
t
w
a
s
p
er
f
o
r
m
ed
b
y
th
e
Dau
g
m
an
r
u
b
b
er
s
h
ee
t
m
o
d
el
w
it
h
a
r
es
o
lu
tio
n
o
f
1
0
×2
4
0
.
T
h
is
s
ta
g
e
w
a
s
a
n
al
y
ze
d
b
y
th
e
b
en
c
h
o
f
t
w
o
1
D
L
o
g
-
Gab
o
r
f
ilter
s
to
ex
tr
ac
t
t
h
e
te
x
tu
r
e
ch
ar
ac
ter
is
tic
s
a
n
d
th
e
en
co
d
in
g
w
as
r
ea
lized
w
it
h
a
p
h
a
s
e
o
f
q
u
an
tizatio
n
d
ev
elo
p
ed
b
y
J
.
Dau
g
m
an
to
g
en
er
ate
t
h
e
b
in
ar
y
ir
i
s
te
m
p
late.
Fo
r
th
e
au
th
e
n
ticatio
n
an
d
th
e
s
i
m
ilar
it
y
m
ea
s
u
r
e
m
e
n
t
b
et
w
ee
n
b
o
t
h
b
in
ar
y
ir
i
s
es
te
m
p
late
s
,
t
h
e
h
a
m
m
i
n
g
d
is
ta
n
ce
s
ar
e
u
s
ed
w
ith
a
p
r
ev
io
u
s
l
y
ca
lcu
lated
t
h
r
es
h
o
ld
.
T
h
e
s
co
r
e
f
u
s
io
n
is
ap
p
lied
u
s
in
g
Dez
er
t
S
m
ar
an
d
ac
h
e
C
la
s
s
ic
(
D
S
m
C
)
r
u
le.
T
h
e
ex
p
er
i
m
en
t
test
ed
o
n
C
a
s
ia
-
ir
i
s
v
3
-
i
n
ter
v
a
l
s
h
o
w
s
t
h
at
th
e
p
r
o
p
o
s
ed
s
y
s
te
m
g
i
v
es
a
g
o
o
d
p
er
f
o
r
m
a
n
c
e
co
m
p
ar
ed
to
o
th
er
s
ap
p
r
o
ac
h
es
w
i
th
a
n
ac
cu
r
ac
y
o
f
9
9
.
9
6
%,
F
A
R
o
f
0
%,
F
R
R
o
f
3
.
8
9
%,
E
E
R
o
f
2
%
an
d
p
r
o
ce
s
s
in
g
t
i
m
e
f
o
r
o
n
e
ir
is
i
m
ag
e
o
f
1
2
.
3
7
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
6
,
Dec
em
b
er
2
0
1
9
:
4
7
0
3
-
4
7
1
2
4712
RE
F
E
R
E
NC
E
S
[1
]
Da
u
g
m
a
n
J.
,
“
Ho
w
iri
s rec
o
g
n
it
io
n
w
o
rk
s
,
”
IEE
E
T
ra
n
s.Circ
u
it
s
S
y
st.
Vi
d
e
o
T
e
c
h
n
,
1
4
:
21
-
30
,
2
0
0
4
.
[2
]
W
il
d
e
s
R.
P
e
t
a
l
.
A
m
a
c
h
in
e
-
v
isio
n
sy
ste
m
f
o
r
iri
s rec
o
g
n
it
io
n
.
M
a
c
h
in
e
Vi
si
o
n
a
n
d
A
p
p
l
ica
ti
o
n
s
.
1
9
9
6
;
9
:
1
-
8
.
[3
]
Bo
les
a
n
d
Bo
a
sh
a
sh
B
.
A
h
u
m
a
n
id
e
n
ti
f
ica
ti
o
n
tec
h
n
iq
u
e
u
sin
g
i
m
a
g
e
s
o
f
th
e
iri
s
a
n
d
w
a
v
e
let
tran
s
f
o
r
m
.
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
S
i
g
n
a
l
Pr
o
c
e
ss
in
g
.
1
9
9
8
;
4
6
:
1
1
8
5
-
1
1
8
8
.
[4
]
Ma
L
e
t
a
l.
“
Eff
i
c
ien
t
iri
s
re
c
o
g
n
it
io
n
b
y
c
h
a
ra
c
teriz
in
g
k
e
y
l
o
c
a
l
v
a
riatio
n
s
,”
IEE
E
T
ra
n
sa
c
t
io
n
s
o
n
Ima
g
e
Pro
c
e
ss
in
g
,
1
3
:
7
3
9
-
7
5
0
,
2
0
0
4
.
[5
]
Ay
d
i
W,
M
a
s
m
o
u
d
i
N,
Ka
m
o
u
n
L.
“
A
F
a
st
a
n
d
A
c
c
u
ra
te
Circu
lar
S
e
g
m
e
n
tatio
n
M
e
t
h
o
d
f
o
r
Iri
s
Re
c
o
g
n
it
io
n
S
y
st
e
m
s
,”
In
ter
n
a
ti
o
n
a
l
Rev
iew o
n
Co
m
p
u
ter
s a
n
d
S
o
ft
wa
re
(
IRE
COS
)
,
9
(3
):
4
6
8
-
4
7
7
,
2
0
1
4
.
[6
]
Biswa
s
R,
Ud
d
in
J,
Ha
sa
n
M
d
.
J.
“
A
Ne
w
A
p
p
ro
a
c
h
o
f
Iris
De
te
c
ti
o
n
a
n
d
Re
c
o
g
n
it
io
n
,”
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
C
o
mp
u
ter
En
g
in
e
e
rin
g
(
IJ
ECE
)
,
7
(
5
):
2
5
3
0
-
2
5
3
6
,
2
0
1
7
.
[7
]
Kh
o
ti
m
a
h
C
a
n
d
Ju
n
iati
D
.
“
Iris
Re
c
o
g
n
it
io
n
Us
in
g
F
e
a
tu
re
Ex
trac
ti
o
n
o
f
Bo
x
Co
u
n
ti
n
g
F
ra
c
ta
l
Di
m
e
n
sio
n
,”
J
.
Ph
y
sic
s: Co
n
f
.
S
e
rie
s
,
2
0
1
8
.
[8
]
Bo
b
e
ld
y
k
D
a
n
d
Ro
ss
A
.
“
Pr
e
d
ictin
g
Eye
Co
lo
r
fro
m
Ne
a
r
In
fra
re
d
Iris
Ima
g
e
s
,”
In
tern
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
Bio
m
e
tri
c
s
,
104
-
1
1
0
,
2
0
1
8
.
[9
]
Ya
n
Z,
He
L
,
Zh
a
n
g
M
,
S
u
n
Z
a
n
d
.
T
a
n
T
.
“
Hi
e
ra
rc
ica
l
M
u
lt
icla
ss
Iris
Cla
ss
if
ica
ti
o
n
fo
r
L
ive
n
e
ss
d
e
tec
ti
o
n
,”
In
tern
a
t
io
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Bi
o
m
e
tri
c
s,
47
-
53
,
2
0
1
8
.
[1
0
]
Da
w
a
y
H.
G
,
k
a
re
e
m
H.
H,
Ha
sh
im
A
.
R
,
“
P
u
p
il
De
tec
ti
o
n
B
a
se
d
o
n
Co
l
o
r
Dif
f
e
re
n
c
e
a
n
d
Circu
lar
Ho
u
g
h
T
ra
n
s
f
o
r
m
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
E
lec
trica
l
a
n
d
Co
mp
u
ter
En
g
i
n
e
e
rin
g
(
IJ
ECE
)
,
8(
5
):
3
2
7
8
-
3
2
8
4
,
2
0
1
8
.
[1
1
]
If
ta
k
h
a
r
Ha
s
a
n
K.M
,
As
h
ra
f
u
l
Am
in
M.
“
Du
a
l
iri
s
m
a
tch
in
g
f
o
r
b
io
m
e
tri
c
id
e
n
ti
f
ica
ti
o
n
,”
S
ig
n
a
l
Ima
g
e
a
n
d
Vi
d
e
o
Pro
c
e
ss
in
g
.
8
(
8
):
1
6
0
5
-
1
6
1
1
,
2
0
1
4
.
[1
2
]
Ng
u
y
e
n
K
,
De
n
m
a
n
S.
“
S
c
o
re
-
L
e
v
e
l
M
u
lt
ib
i
o
m
e
tri
c
F
u
sio
n
Ba
se
d
o
n
De
m
p
ste
r
-
S
h
a
f
e
r
T
h
e
o
r
y
In
c
o
rp
o
ra
ti
n
g
Un
c
e
rtain
ty
F
a
c
to
rs
,”
IEE
E
T
ra
n
sa
c
ti
o
n
s o
n
Hu
ma
n
-
M
a
c
h
i
n
e
S
y
ste
ms
,
4
5
:
1
3
2
-
1
4
0
,
2
0
1
5
.
[1
3
]
Dw
i
v
e
d
i
R,
De
y
S.
“
S
c
o
re
lev
e
l
f
u
sio
n
f
o
r
c
o
n
c
e
lab
le
m
u
lt
i
-
b
io
m
e
tri
c
v
e
ri
f
ic
a
ti
o
n
,”
Pa
tt
e
rn
Rec
o
g
n
it
io
n
L
e
tt
e
rs
,
2
0
1
8
.
[1
4
]
G
h
a
l
e
m
K
.
G
,
H
e
n
d
e
l
F
.
“
A
u
th
e
n
t
i
c
a
t
i
o
n
a
n
d
id
e
n
ti
fi
c
a
ti
o
n
o
f
in
di
v
i
d
u
a
l
s
fr
o
m
th
e
i
r
i
s
i
ma
g
e
s
,”
S
e
c
on
d
I
nt
e
r
n
a
ti
o
n
a
l
W
or
k
s
h
o
p
o
n
M
a
th
e
m
a
t
i
cs
a
n
d
C
o
m
p
u
t
e
r
S
c
i
e
n
ce
(
IW
MC
S)
,
2
01
4
.
[1
5
]
S
m
a
ra
n
d
a
c
h
e
F
a
n
d
De
z
e
rt
J.
“
A
p
p
li
c
a
ti
o
n
s
a
n
d
A
d
v
a
n
c
e
s
o
f
DSmT
f
o
r
In
f
o
r
m
a
ti
o
n
F
u
si
o
n
(Co
ll
e
c
ted
w
o
rk
s)
,”
1
,
2
0
0
4
.
[1
6
]
S
m
a
r
a
n
d
a
c
h
e
F
a
n
d
De
z
e
rt
J.
“
A
p
p
li
c
a
ti
o
n
s
a
n
d
A
d
v
a
n
c
e
s
o
f
D
S
m
T
f
o
r
In
f
o
rm
a
ti
o
n
F
u
si
o
n
(Co
ll
e
c
ted
w
o
rk
s)
,”
2
,
2
0
0
4
.
[1
7
]
S
m
a
r
a
n
d
a
c
h
e
F
e
t
De
z
e
rt
J.
“
Ap
p
li
c
a
ti
o
n
s
a
n
d
A
d
v
a
n
c
e
s
o
f
DS
m
T
f
o
r
In
f
o
rm
a
ti
o
n
F
u
sio
n
(Co
l
lec
ted
w
o
rk
s)
,”
3
,
2
0
0
9
.
[1
8
]
Iris
d
a
tab
a
se
CA
S
I
A
-
IrisV
3
,
“
Ch
i
n
e
se
A
c
a
d
e
m
y
o
f
S
c
ien
c
e
s
-
In
stit
u
te o
f
A
u
to
m
a
ti
o
n
,”
Re
tri
e
v
e
d
o
n
De
c
2
0
1
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.