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
.
5
,
Octo
b
e
r
2
0
1
7
,
p
p
.
2
5
30
~
2
536
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
7
i
5
.
pp
2
5
30
-
2
536
2530
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
A New
Approa
ch
o
f
Iris
De
tect
io
n
and Reco
g
nition
Rubel B
is
w
a
s
,
J
ia
Uddi
n
,
M
d.
J
un
a
y
ed
H
a
s
a
n
De
p
a
rte
m
e
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
E
n
g
in
e
e
rin
g
,
BRA
C
Un
iv
e
rsity
,
Ba
n
g
lad
e
sh
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
2
6
,
2
0
1
7
R
ev
i
s
ed
J
u
n
1
7
,
2
0
1
7
A
cc
ep
ted
Sep
11
,
2
0
1
7
T
h
is
p
a
p
e
r
p
ro
p
o
se
s
a
n
IRIS
re
c
o
g
n
it
io
n
a
n
d
d
e
tec
ti
o
n
m
o
d
e
l
f
o
r
m
e
a
su
rin
g
th
e
e
-
se
c
u
rit
y
.
T
h
is
p
ro
p
o
se
d
m
o
d
e
l
c
o
n
sists
o
f
th
e
f
o
ll
o
w
in
g
b
lo
c
k
s:
se
g
m
e
n
tatio
n
a
n
d
n
o
rm
a
li
z
a
ti
o
n
,
f
e
a
tu
re
e
n
c
o
d
in
g
a
n
d
f
e
a
tu
re
e
x
tr
a
c
ti
o
n
,
a
n
d
c
las
si
f
ica
ti
o
n
.
In
f
irst
p
h
a
se
,
h
is
to
g
ra
m
e
q
u
a
li
z
a
ti
o
n
a
n
d
c
a
n
n
y
e
d
g
e
d
e
tec
ti
o
n
is
u
se
d
f
o
r
o
b
jec
t
d
e
tec
ti
o
n
.
A
n
d
th
e
n
,
Ho
u
g
h
T
ra
n
sf
o
rm
a
ti
o
n
is
u
ti
li
z
e
d
f
o
r
d
e
tec
ti
n
g
th
e
c
e
n
ter
o
f
th
e
p
u
p
il
o
f
a
n
IRIS
.
In
se
c
o
n
d
p
h
a
se
,
Da
u
g
m
e
n
’s
Ru
b
b
e
r
S
h
e
e
t
m
o
d
e
l
a
n
d
L
o
g
G
a
b
o
r
f
il
ter
is
u
se
d
f
o
r
n
o
rm
a
li
z
a
ti
o
n
a
n
d
e
n
c
o
d
i
n
g
a
n
d
a
s
a
f
e
a
tu
re
e
x
tra
c
ti
o
n
m
e
th
o
d
G
NS
(
G
lo
b
a
l
Ne
ig
h
b
o
r
h
o
o
d
S
tru
c
tu
re
)
m
a
p
is
u
se
d
,
f
in
a
ll
y
e
x
trac
ted
f
e
a
tu
re
o
f
G
NS
is
fe
e
d
to
th
e
S
VM
(S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
)
f
o
r
tr
a
in
in
g
a
n
d
tes
ti
n
g
.
F
o
r
o
u
r
tes
ted
d
a
tas
e
t,
e
x
p
e
ri
m
e
n
tal
re
su
lt
s
d
e
m
o
n
s
trate
9
2
%
a
c
c
u
ra
c
y
in
re
a
l
p
o
rti
o
n
a
n
d
8
6
%
a
c
c
u
ra
c
y
in
i
m
a
g
in
a
r
y
p
o
rti
o
n
f
o
r
b
o
th
e
y
e
s.
In
a
d
d
it
io
n
,
o
u
r
p
ro
p
o
se
d
m
o
d
e
l
o
u
t
p
e
rf
o
rm
s
th
a
n
o
t
h
e
r
tw
o
c
o
n
v
e
n
ti
o
n
a
l
m
e
th
o
d
s
e
x
h
ib
i
ti
n
g
h
ig
h
e
r
a
c
c
u
ra
c
y
.
K
ey
w
o
r
d
s
:
Dau
g
m
a
n
’
s
r
u
b
b
er
s
h
ee
t
DNS
(
Do
m
i
n
an
t N
eig
h
b
o
r
h
o
o
d
Stru
ct
u
r
e)
E
-
Sec
u
ir
it
y
G
NS
(
G
lo
b
a
l
Ne
ig
h
b
o
r
h
o
o
d
S
tru
c
tu
re
)
I
r
is
r
ec
o
g
n
itio
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
:
J
ia
Ud
d
in
,
A
s
s
is
ta
n
t P
r
o
f
ess
o
r
Dep
ar
te
m
en
t
o
f
C
o
m
p
u
ter
Sci
en
ce
an
d
E
n
g
i
n
ee
r
in
g
,
B
R
AC
U
n
iv
er
s
it
y
,
6
6
Mo
h
ak
h
ali,
D
h
ak
a
1
2
1
2
,
B
an
g
lad
e
s
h
.
E
m
ail: j
ia.
u
d
d
in
@
b
r
ac
u
.
ac
.
b
d
1.
I
NT
RO
D
UCT
I
O
N
Fo
r
th
e
a
u
th
e
n
tica
tio
n
o
f
t
h
e
id
en
tit
y
o
f
a
n
i
n
d
i
v
id
u
al,
b
io
m
etr
ic
r
ec
o
g
n
itio
n
is
a
r
eliab
l
e
w
a
y
.
Fo
r
b
io
m
etr
ic
a
u
th
e
n
ticat
io
n
,
s
ev
e
r
al
p
h
y
s
ical
s
tab
le
ch
ar
ac
ter
is
t
ics,
s
u
c
h
a
s
f
i
n
g
er
p
r
in
t
s
,
v
o
ic
e
r
ec
o
g
n
i
tio
n
,
h
a
n
d
g
eo
m
etr
y
,
h
a
n
d
w
r
iti
n
g
,
t
h
e
r
e
tin
a,
ir
is
r
ec
o
g
n
it
io
n
ar
e
u
s
ed
.
Am
o
n
g
t
h
e
c
h
ar
ac
ter
is
tic
s
,
m
o
s
t
o
f
t
h
e
m
r
eq
u
ir
e
th
e
s
o
m
e
p
h
y
s
ical
ac
tio
n
s
o
f
s
en
s
in
g
d
e
v
ic
e
o
r
s
o
m
e
s
p
ec
i
al
ac
tio
n
s
[
1
]
.
T
o
o
v
er
co
m
e
t
h
e
b
ar
r
ier
lik
e
th
e
p
h
y
s
ical
co
n
tac
ts
,
a
u
to
m
a
ted
r
ec
o
g
n
itio
n
tec
h
n
iq
u
e
i
s
a
n
alter
n
ati
v
e
a
n
d
les
s
i
n
v
asi
v
e.
I
r
is
r
ec
o
g
n
i
tio
n
tech
n
iq
u
e
is
o
n
e
o
f
th
e
m
[
1
3
]
.
Fig
u
r
e
1
.
Deta
ils
o
f
t
h
e
p
o
r
tio
n
s
o
f
h
u
m
a
n
e
y
e
T
h
is
n
o
n
i
n
v
asi
v
e
v
er
if
ica
tio
n
tech
n
iq
u
e
f
o
r
t
h
e
id
en
ti
f
i
ca
tio
n
o
f
i
n
d
iv
id
u
als
is
m
o
r
e
p
r
ac
tical
b
ec
au
s
e
t
h
e
p
atter
n
o
f
th
e
h
u
m
an
ir
is
is
u
n
iq
u
e,
d
i
s
ti
n
ctiv
e
an
d
s
tab
le
th
r
o
u
g
h
o
u
t
th
e
ad
u
lt
l
if
e
[
1
]
.
Fig
u
r
e
1
illu
s
tr
ates
a
g
e
n
er
alize
d
o
v
er
v
ie
w
o
f
h
u
m
a
n
e
y
e;
f
r
o
m
i
t
w
e
ca
n
s
ee
th
e
p
o
s
itio
n
o
f
ea
ch
p
o
r
tio
n
s
o
f
th
e
e
y
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
CE
I
SS
N:
2088
-
8708
A
N
ew A
p
p
r
o
a
ch
o
f I
r
is
Dete
ctio
n
a
n
d
R
ec
o
g
n
itio
n
(
R
u
b
el
B
is
w
a
s
)
2531
Fro
m
t
h
e
h
u
m
a
n
e
y
e
s
,
w
e
ca
n
g
et
t
h
e
p
atter
n
o
f
t
h
e
i
r
is
w
h
ich
is
id
en
tical
a
n
d
r
eliab
le
f
o
r
h
u
m
a
n
id
en
ti
f
icatio
n
p
u
r
p
o
s
e.
P
r
ev
io
u
s
l
y
,
a
n
u
m
b
er
o
f
r
esear
c
h
er
s
w
o
r
k
ed
o
n
I
r
is
r
ec
o
g
n
iti
o
n
.
I
n
m
o
s
t
o
f
th
e
co
n
v
e
n
tio
n
al
m
et
h
o
d
s
,
r
esear
ch
er
s
w
o
r
k
ed
o
n
eit
h
er
I
r
is
lo
ca
lizatio
n
o
r
I
r
is
p
atter
n
r
e
co
g
n
itio
n
.
Fo
r
I
r
is
lo
ca
lizatio
n
m
o
s
t
o
f
t
h
e
r
esea
r
ch
er
s
u
s
e
Da
u
g
m
a
n
’
s
I
r
is
r
ec
o
g
n
itio
n
m
o
d
el
[
2
]
.
I
n
Dau
g
m
an
’
s
m
o
d
el,
a
2
5
6
b
y
te
co
d
e
is
g
e
n
er
ated
b
y
q
u
a
n
tizi
n
g
th
e
lo
ca
l
p
h
ase
a
n
g
le
b
ased
o
n
th
e
r
es
u
lt
o
f
th
e
f
ilt
er
ed
im
a
g
e.
W
ith
a
q
u
iet
d
if
f
er
e
n
t
ap
p
r
o
ac
h
th
e
W
ild
es
s
y
s
te
m
f
r
ag
m
e
n
ted
a
s
tan
d
ar
d
ized
co
n
n
ec
tio
n
b
as
ed
o
n
g
o
o
d
n
es
s
-
of
-
m
atc
h
v
al
u
es
a
n
d
Fis
h
er
’
s
li
n
e
ar
d
is
cr
i
m
in
a
n
t
f
o
r
p
atter
n
id
en
ti
f
icatio
n
[
3
]
.
J
iali
C
u
i
el
a
l.
[
4
]
p
r
o
p
o
s
ed
an
ir
is
d
etec
tio
n
alg
o
r
it
h
m
b
ased
o
n
p
r
in
cip
al
co
m
p
o
n
e
n
t
an
a
l
y
s
is
.
H
y
u
n
g
G
u
L
ee
el
a
l.
u
tili
ze
d
th
e
b
i
n
ar
y
f
ea
t
u
r
e
f
o
r
I
r
is
d
etec
tio
n
[
5
]
.
On
th
e
s
e
v
ar
ietie
s
o
f
m
et
h
o
d
,
f
ea
t
u
r
e
v
ec
to
r
is
n
o
t
m
ai
n
k
e
y
f
ea
tu
r
e
f
o
r
d
etec
tio
n
tech
n
iq
u
es
ex
ce
p
t
th
e
Dau
g
m
an
’
s
m
o
d
el.
Fo
r
th
e
n
u
m
er
ica
l
p
r
esen
tatio
n
o
f
th
e
ch
ar
ac
ter
o
f
an
y
o
b
j
ec
t,
f
o
r
th
e
b
en
ef
it
i
n
s
tatis
t
ical
p
r
o
ce
d
u
r
es
an
d
d
ef
i
n
in
g
t
h
e
ch
ar
ac
ter
is
tics
i
n
to
s
o
m
e
d
i
m
e
n
s
io
n
al
v
ec
to
r
s
,
f
ea
t
u
r
e
v
ec
to
r
ap
p
r
o
ac
h
is
m
o
r
e
r
eliab
le.
B
esid
e
th
ese,
s
el
ec
ti
n
g
t
h
e
o
p
ti
m
al
f
ea
t
u
r
e
is
also
a
v
ital c
r
iter
ia.
I
n
o
u
r
p
r
o
p
o
s
ed
m
o
d
el,
w
e
u
t
ilized
th
e
Da
u
g
m
a
n
’
s
m
o
d
el.
Af
ter
f
ea
t
u
r
e
e
n
co
d
in
g
,
DNS
an
d
GNS
m
ap
s
[
6
,
7
]
ar
e
u
s
ed
to
r
ed
u
c
e
th
e
s
ize
o
f
t
h
e
f
ea
tu
r
e
v
ec
to
r
an
d
ex
tr
ac
t
th
e
f
ea
t
u
r
e
v
ec
to
r
o
f
o
u
r
test
ed
I
R
I
S
d
ataset.
Fo
r
th
e
m
ac
h
i
n
e
lear
n
in
g
p
u
r
p
o
s
e
o
n
r
ec
o
g
n
i
tio
n
p
h
ase,
s
in
g
le
cla
s
s
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
e
h
as
b
ee
n
u
s
ed
.
I
n
ad
d
itio
n
,
w
e
h
av
e
co
m
p
ar
ed
th
e
p
er
f
o
r
m
a
n
ce
o
f
o
u
r
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
co
n
s
id
er
in
g
th
e
s
a
m
e
en
v
ir
o
n
m
e
n
t
w
i
th
t
w
o
o
t
h
er
ap
p
r
o
ac
h
es:
o
n
e
is
p
r
o
p
o
s
ed
b
y
L
i
M.
el
a
l.
[
8
]
,
an
o
th
er
o
n
e
b
y
K.
Sath
i
y
ar
aj
a
el
a
l.
[
9
]
.
H
o
w
ev
er
,
th
e
r
ec
en
t
p
e
r
f
o
r
m
an
ce
co
m
p
a
r
is
o
n
in
th
e
ar
ea
o
f
i
r
is
r
ec
o
g
n
iti
o
n
an
d
d
e
tect
io
n
d
e
p
en
d
s
o
n
h
o
w
f
ar
th
e
a
cc
y
r
ac
y
,
ef
f
icac
y
an
d
s
ca
la
b
il
ity
p
e
r
f
o
r
m
an
ce
ca
n
b
e
in
cr
ea
s
e
d
[
1
5
]
.
I
r
is
d
ec
tio
n
an
d
r
ec
o
g
n
itio
n
i
s
an
i
m
p
o
r
tan
t
p
r
e
-
p
r
o
ce
s
s
in
g
s
tep
in
au
to
m
a
tic
au
to
m
atic
s
y
s
te
m
s
an
d
a
w
ell
-
d
esi
g
n
ed
tech
n
iq
u
e
ca
n
i
m
p
r
o
v
e
th
e
ac
c
u
r
ac
y
i
n
co
llectin
g
clea
r
ir
is
i
m
a
g
es
a
n
d
m
ar
k
n
o
is
e
ar
ea
s
[
1
6
]
.
T
h
e
p
ap
er
is
o
r
g
an
ized
a
s
f
o
llo
w
s
–
Sect
io
n
2
d
escr
ib
es
t
h
e
o
v
er
v
ie
w
i
n
d
ep
th
o
f
o
u
r
p
r
o
p
o
s
ed
m
o
d
el.
T
h
e
r
esu
lts
o
f
t
h
e
ex
p
er
i
m
en
t
s
ca
r
r
ied
o
u
t
an
d
th
eir
an
al
y
s
i
s
ar
e
in
clu
d
ed
in
Sec
tio
n
3
an
d
f
in
all
y
Sectio
n
4
co
n
clu
d
es t
h
e
p
ap
er
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
Fig
u
r
e
2
p
r
esen
ts
a
d
etailed
i
m
p
le
m
e
n
tat
io
n
o
f
o
u
r
r
esear
c
h
m
o
d
el
,
in
w
h
ic
h
w
e
h
ig
h
li
g
h
t
th
e
m
aj
o
r
p
o
r
tio
n
s
b
y
d
r
a
w
i
n
g
s
ep
er
ate
b
lo
ck
s
.
Fig
u
r
e
2
.
T
h
e
f
u
ll st
ep
s
o
f
th
e
I
R
I
S d
etec
tio
n
an
d
r
ec
o
g
n
itio
n
m
o
d
el
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
.
5
,
Octo
b
er
201
7
:
25
30
–
2
5
3
6
2532
2
.
1
.
S
eg
m
e
nta
t
io
n
A
t
t
h
e
b
eg
i
n
n
in
g
o
f
p
r
o
ce
s
s
in
g
o
f
th
e
i
n
p
u
t
i
m
a
g
e,
s
o
m
e
s
tep
s
ar
e
r
eq
u
ir
ed
to
en
s
u
r
e
b
etter
p
er
f
o
r
m
a
n
ce
f
r
o
m
t
h
e
s
y
s
te
m
.
W
e
h
a
v
e
u
s
ed
h
is
to
g
r
a
m
eq
u
aliza
tio
n
tec
h
n
iq
u
e
to
ad
j
u
s
t
t
h
e
i
m
a
g
e
i
n
te
n
s
it
ie
s
in
o
r
d
er
to
en
h
an
ce
co
n
tr
a
s
t.
Fo
r
im
p
r
o
v
i
n
g
th
e
ed
g
e
d
ete
ctio
n
,
w
e
h
av
e
co
n
s
id
er
ed
im
a
g
e
ad
j
u
s
t
m
e
n
t.
Fo
r
th
e
ed
g
e
d
etec
tio
n
p
ar
t,
w
e
h
a
v
e
u
s
ed
C
a
n
n
y
E
d
g
e
Dete
c
tio
n
A
l
g
o
r
it
h
m
(
C
E
D
A
)
[
1
0
]
.
C
E
DA
is
a
m
u
lti
-
s
tag
e
alg
o
r
ith
m
to
d
etec
t
a
w
id
e
r
a
n
g
e
o
f
ed
g
e
s
i
n
t
h
e
i
m
a
g
e.
B
y
s
m
o
o
th
i
n
g
t
h
e
i
m
a
g
e
w
it
h
t
h
e
h
elp
o
f
Ga
u
s
s
ia
n
f
ilter
t
h
e
n
o
is
e
w
il
l
b
e
r
e
m
o
v
e
d
.
T
h
en
t
h
is
a
lg
o
r
it
h
m
f
i
n
d
s
t
h
e
i
n
ten
s
it
y
g
r
ad
ien
ts
o
f
t
h
e
i
m
ag
e
an
d
t
h
en
n
o
n
-
m
ax
i
m
u
m
s
u
p
p
r
ess
io
n
is
ap
p
l
ied
to
g
et
r
id
o
f
s
p
u
r
io
u
s
r
es
p
o
n
s
e
to
ed
g
e
d
etec
tio
n
.
Fig
u
r
e
2
s
h
o
w
s
t
h
at
in
Seg
m
en
tatio
n
p
o
r
tio
n
(
A
)
,
f
o
r
p
u
p
il
an
d
b
o
u
n
d
ar
y
s
elec
ti
o
n
Ho
g
h
tr
an
s
f
o
r
m
atio
n
h
as
b
ee
n
u
s
ed
an
d
f
o
r
s
cler
a/I
R
I
S d
etec
tio
n
,
m
id
p
o
in
t a
lg
o
r
it
h
m
h
a
s
b
ee
n
u
s
ed
.
Fo
r
ir
is
d
etec
tio
n
,
in
o
u
r
s
y
s
te
m
,
w
e
h
av
e
i
m
p
le
m
e
n
ted
an
a
u
to
m
at
ic
s
eg
m
e
n
tatio
n
p
r
o
ce
s
s
w
it
h
t
h
e
h
elp
o
f
t
w
o
al
g
o
r
ith
m
s
.
I
n
itia
l
l
y
,
w
e
c
h
ec
k
f
o
r
t
h
e
co
r
n
ea
l
r
ef
lectio
n
.
W
e
h
a
v
e
tak
e
n
t
h
e
co
m
p
le
m
e
n
t
o
f
t
h
e
g
iv
e
n
i
m
a
g
e
a
n
d
th
e
n
w
e
r
em
o
v
ed
t
h
e
d
ar
k
p
o
in
ts
(
r
ef
lec
tio
n
p
o
in
t
s
)
.
A
f
ter
th
at,
f
r
o
m
th
e
g
iv
e
n
i
m
a
g
e
w
e
h
av
e
ca
lc
u
lated
t
h
e
f
ir
s
t
d
er
iv
ativ
es
o
f
i
n
ten
s
it
y
v
al
u
e
s
b
y
c
alcu
lati
n
g
th
e
r
es
u
l
t
b
ased
o
n
th
r
es
h
o
ld
v
al
u
e
w
e
g
en
er
ate
a
n
ed
g
e
m
ap
.
T
h
e
p
a
r
a
m
eter
s
o
f
cir
cles
(
ce
n
ter
co
o
r
d
in
ates
an
d
th
e
r
ad
iu
s
)
ar
e
ev
alu
a
ted
b
y
v
o
ti
n
g
in
Ho
u
g
h
s
p
ac
e.
I
n
o
r
d
er
to
d
etec
t
th
e
p
u
p
il
w
e
b
ias
t
h
e
f
ir
s
t
d
er
iv
ativ
e
in
v
er
tical
d
ir
ec
tio
n
.
Fi
n
all
y
,
w
e
h
av
e
dr
a
w
n
a
cir
cle
b
y
d
o
u
b
li
n
g
t
h
e
p
u
p
il r
ad
iu
s
.
2
.
2
.
No
r
m
a
liza
t
io
n
Dau
g
m
a
n
’
s
m
o
d
el
[
2
]
is
u
s
ed
f
o
r
t
h
e
n
o
r
m
aliza
tio
n
o
f
o
u
r
s
eg
m
e
n
ted
ir
is
r
e
g
io
n
s
.
W
e
h
av
e
m
ad
e
s
u
r
e
t
h
at
f
o
r
all
th
e
n
o
r
m
alize
d
im
a
g
e
s
m
u
s
t
h
a
v
e
th
e
s
a
m
e
r
eso
lu
tio
n
.
W
e
co
n
s
id
er
ed
th
e
ce
n
ter
o
f
th
e
p
u
p
i
l
as
th
e
r
ef
er
en
ce
p
o
in
t.
W
e
h
av
e
p
ass
ed
th
e
r
ad
ial
v
ec
to
r
s
th
r
o
u
g
h
th
e
ir
i
s
r
eg
io
n
.
T
h
e
s
elec
ted
d
ata
p
o
in
ts
alo
n
g
ea
ch
r
ad
ial
lin
e
ar
e
k
n
o
w
n
a
s
r
ad
ial
r
eso
lu
tio
n
.
T
h
e
n
u
m
b
er
o
f
r
ad
ial
lin
es
g
o
i
n
g
ar
o
u
n
d
th
e
ir
is
r
eg
io
n
ar
e
k
n
o
w
n
a
s
r
ad
ial
r
eso
lu
tio
n
w
h
er
e
t
h
e
n
u
m
b
er
o
f
r
ad
ial
lin
es
g
o
in
g
ar
o
u
n
d
t
h
e
ir
is
r
e
g
io
n
ar
e
d
ef
i
n
ed
as
an
g
u
lar
r
eso
lu
tio
n
[
1
4
]
.
B
ec
au
s
e
o
f
t
h
e
n
o
n
-
co
n
ce
n
tr
ic
n
atu
r
e
o
f
th
e
p
u
p
il to
th
e
ir
is
,
to
r
escale
p
o
in
ts
a
r
e
m
ap
p
in
g
f
o
r
m
u
la
i
s
n
ee
d
ed
b
ased
o
n
th
e
an
g
le
ar
o
u
n
d
t
h
e
cir
cle.
T
h
e
f
o
r
m
u
la
i
s
(1
)
W
ith
(2
)
(
3
)
Her
e,
(
O
x
, O
y
)
r
ep
r
esen
ts
t
h
e
ce
n
ter
d
is
p
lace
m
e
n
t o
f
t
h
e
p
u
p
i
l c
o
m
p
ar
e
to
ir
is
ce
n
ter
.
r
′
r
ep
r
esen
t
s
th
e
d
is
tan
ce
b
et
w
ee
n
th
e
ed
g
e
s
o
f
p
u
p
il a
n
d
ir
is
.
is
t
h
e
an
g
le
b
as
ed
o
n
th
e
ed
g
es
w
er
e
co
u
n
ted
.
r
1
is
th
e
r
ad
iu
s
o
f
th
e
ir
is
[
1
4
]
.
Fig
u
r
e
3
.
R
ec
tan
g
u
lar
s
h
ee
t f
o
r
m
atio
n
f
r
o
m
cir
c
u
lar
ar
ea
o
f
Dau
g
m
a
n
’
s
r
u
b
b
er
s
h
ee
t
m
o
d
el
T
h
e
f
u
n
ctio
n
f
ir
s
t
g
i
v
es
a
‘
d
o
u
g
h
n
u
t
’
f
o
r
m
to
t
h
e
ir
i
s
r
eg
io
n
b
ased
o
n
th
e
an
g
le.
Fro
m
t
h
is
‘
d
o
u
g
h
n
u
t’
f
o
r
m
ir
is
r
e
g
io
n
w
e
co
n
s
tr
u
ct
a
2
D
ar
r
a
y
w
it
h
h
o
r
izo
n
tal
d
i
m
en
s
io
n
o
f
a
n
g
u
lar
r
eso
l
u
tio
n
a
n
d
v
er
tical
d
i
m
e
n
s
io
n
o
f
th
e
r
ad
ial
r
eso
lu
tio
n
[
1
]
.
Fig
u
r
e
3
s
h
o
w
s
t
h
e
m
a
in
s
tr
u
ct
u
r
al
v
is
u
ali
za
tio
n
o
f
p
r
o
d
u
cin
g
th
e
r
ec
tan
g
u
lar
ar
ea
f
r
o
m
th
e
c
ir
cu
lar
r
ad
iu
s
o
f
Da
u
g
m
a
n
’
s
r
u
b
b
er
s
h
ee
t
m
o
d
el.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
CE
I
SS
N:
2088
-
8708
A
N
ew A
p
p
r
o
a
ch
o
f I
r
is
Dete
ctio
n
a
n
d
R
ec
o
g
n
itio
n
(
R
u
b
el
B
is
w
a
s
)
2533
2
.
3
.
F
ea
t
ure
E
nco
din
g
T
h
e
te
m
p
late
m
atr
ix
s
ize
i
s
s
et
in
s
u
c
h
a
w
a
y
b
y
d
o
u
b
li
n
g
th
e
co
l
u
m
n
s
ize
o
f
th
e
n
o
r
m
alize
d
ir
is
i
m
a
g
e
an
d
t
h
e
r
o
w
i
s
k
ep
t
s
a
m
e.
T
h
e
r
ea
s
o
n
b
e
h
in
d
t
h
i
s
–
in
o
u
r
te
m
p
late
b
o
th
th
e
r
ea
l
an
d
i
m
a
g
i
n
ar
y
v
al
u
e
w
il
l
p
ar
ticip
ate.
Af
ter
t
h
at,
t
h
e
n
o
r
m
al
ize
d
ir
is
p
atter
n
i
s
co
n
v
o
l
u
ted
w
it
h
1
D
L
o
g
-
Gab
o
r
w
a
v
elet
s
.
Firs
t
1
D
s
ig
n
al
s
ar
e
g
e
n
er
ated
f
r
o
m
2
D
n
o
r
m
alize
d
ir
i
s
p
atter
n
a
n
d
th
en
Gab
o
r
f
ilter
i
s
u
s
ed
to
th
o
s
e
1
D
s
ig
n
al
s
.
I
n
t
h
e
L
o
g
–
Gab
o
r
eq
u
atio
n
w
e
u
s
ed
th
e
f
o
llo
w
in
g
v
al
u
e.
T
h
e
v
al
u
e
f
0
is
s
et
to
1
8
w
h
ic
h
r
ep
r
ese
n
ts
a
s
ca
le
4
Gab
o
r
w
a
v
elet.
Fro
m
t
h
e
e
x
p
er
i
m
e
n
t,
w
e
s
et
t
h
e
v
alu
e
o
f
s
i
g
m
a
o
v
e
r
f
r
eq
u
en
c
y
to
th
e
0
.
5
.
(
4
)
W
h
er
e
G
is
t
h
e
Gab
o
r
f
ilter
e
d
f
u
n
ctio
n
.
f
0
an
d
σ
ar
e
th
e
p
ar
am
eter
s
o
f
t
h
e
f
ilter
.
f
0
w
i
ll
g
i
v
e
t
h
e
ce
n
ter
f
r
eq
u
e
n
c
y
o
f
t
h
e
f
ilter
[
1
2
]
.
L
o
g
-
Gab
o
r
f
ilter
r
etu
r
n
s
a
m
a
tr
ix
w
it
h
co
m
p
lex
v
alu
e
d
elem
e
n
t
w
it
h
th
e
s
ize
o
f
n
o
r
m
alize
d
ir
i
s
i
m
a
g
e.
Af
ter
t
h
at,
t
w
o
n
e
w
m
atr
i
x
is
cr
ea
ted
f
r
o
m
b
ased
o
n
th
e
r
ea
l
p
ar
t a
n
d
i
m
a
g
i
n
ar
y
p
ar
t
o
f
L
o
g
-
Gab
o
r
r
etu
r
n
ed
m
atr
ix
.
T
h
er
ea
f
ter
,
t
h
e
r
a
w
d
ata
o
f
t
h
e
s
e
i
s
co
n
v
er
ted
to
p
s
e
u
d
o
-
p
o
lar
co
o
r
d
in
ate
s
y
s
te
m
.
T
h
e
n
,
t
h
e
v
alu
e
s
o
f
t
h
e
r
ea
l
p
ar
t
m
atr
i
x
a
n
d
i
m
a
g
in
ar
y
p
ar
t
m
a
tr
ix
in
co
n
v
er
ted
in
to
b
in
ar
y
v
al
u
e.
Fin
all
y
,
b
y
m
er
g
i
n
g
t
h
e
s
e
t
w
o
m
atr
i
x
w
e
g
et
t
h
e
te
m
p
late
o
f
a
h
u
m
a
n
.
T
h
e
d
ata
is
s
et
in
s
u
c
h
a
w
a
y
t
h
at
t
h
e
o
d
d
c
o
lu
m
n
s
co
n
tai
n
s
r
ea
l
p
ar
t
m
a
tr
ix
v
alu
e
a
n
d
th
e
ev
e
n
co
lu
m
n
s
h
o
ld
s
i
m
a
g
i
n
ar
y
m
at
r
ix
v
al
u
e.
W
e
also
ca
lcu
lated
ab
s
o
l
u
te
v
al
u
es.
Fi
g
u
r
e
4
s
h
o
w
s
th
e
p
h
ase
q
u
a
n
tizatio
n
p
r
o
ce
s
s
in
s
h
o
r
t.
Fr
o
m
t
h
is
f
ig
u
r
e,
t
h
e
p
r
o
ce
s
s
o
f
co
llectin
g
th
e
r
ea
l
an
d
i
m
a
g
i
n
ar
y
r
esp
o
n
s
e
o
f
th
e
i
m
ag
e
a
f
ter
ap
p
l
y
i
n
g
ap
p
l
y
i
n
g
L
o
g
Gab
o
r
Fil
ter
;
h
as b
ee
n
v
is
u
alize
d
.
Fig
u
r
e
4
.
P
h
ase
q
u
an
tizat
io
n
p
r
o
ce
s
s
o
f
f
ea
t
u
r
e
en
co
d
in
g
2
.
4
.
T
ex
t
ure
F
ea
t
ure
E
x
t
a
rc
t
io
n
Fo
r
f
ea
tu
r
e
e
x
tr
ac
tio
n
,
f
r
o
m
th
e
Gab
o
r
f
ilter
ed
i
m
ag
e
g
et
tin
g
f
r
o
m
f
ea
t
u
r
e
en
co
d
i
n
g
p
h
ase,
w
e
co
n
s
tr
u
ct
t
h
e
DN
S
[
6
]
m
ap
s
a
cc
o
r
d
in
g
to
t
h
e
s
tati
s
tical
p
ar
am
eter
s
o
f
T
ab
le
1
.
Fro
m
T
ab
le
1
w
e
ar
e
g
etti
n
g
th
e
ex
ac
t
p
ar
a
m
eter
s
o
f
th
e
GNS
an
d
DNS
m
ap
.
W
e
th
en
ca
lcu
late
G
NS
f
ea
t
u
r
e
v
ec
t
o
r
s
b
y
av
er
a
g
in
g
t
h
e
DNS
[
6
,
7
]
v
alu
e
s
o
f
t
h
e
f
ilter
ed
i
m
ag
e.
Fo
r
s
elec
ti
n
g
t
h
e
m
o
s
t
s
i
g
n
if
ica
n
t
f
ea
t
u
r
es
f
r
o
m
t
h
e
GN
S
m
ap
w
h
ic
h
ex
h
ib
it
s
s
p
atial
te
x
tu
r
e
s
,
w
e
s
elec
t
o
n
l
y
t
h
e
f
ea
t
u
r
es
in
t
h
e
co
n
ce
n
tr
ic
cir
cles
o
f
v
ar
io
u
s
r
ad
ii
at
th
e
ce
n
ter
o
f
th
e
m
ap
.
I
n
o
u
r
ex
p
er
i
m
e
n
t,
8
an
d
1
6
f
ea
tu
r
es
f
r
o
m
f
ir
s
t
t
wo
in
n
er
m
o
s
t
cir
cles,
2
4
u
n
if
o
r
m
a
n
g
u
lar
f
ea
tu
r
e
s
f
r
o
m
ea
ch
o
f
th
e
o
t
h
er
8
cir
cles
ar
e
co
n
s
id
er
ed
to
co
n
s
tr
u
ct
th
e
f
ea
t
u
r
e
v
ec
to
r
.
T
h
er
ef
o
r
e,
th
e
n
u
m
b
er
o
f
d
i
m
en
s
io
n
s
o
f
t
h
e
f
ea
t
u
r
e
v
ec
to
r
is
2
1
6
(
=8
+1
6
+8
x
2
4
)
.
Fig
u
r
e
6
ex
h
ib
it
s
th
e
s
tep
s
o
f
G
NS
an
d
DNS
m
ap
ex
tr
ac
tio
n
f
r
o
m
Gab
o
r
f
ilter
ed
i
m
ag
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
.
5
,
Octo
b
er
201
7
:
25
30
–
2
5
3
6
2534
T
ab
le
1
.
Statis
tical
p
ar
a
m
eter
s
o
f
DNS
an
d
GN
S
m
ap
s
u
s
ed
i
n
th
e
e
x
p
er
i
m
e
n
t
Pa
r
a
m
e
t
e
r
s
V
a
l
u
e
s
S
e
a
r
c
h
i
n
g
w
i
n
d
o
w
si
z
e
2
1
x
2
1
N
e
i
g
h
b
o
r
w
i
n
d
o
w
1
3
x
1
3
N
u
mb
e
r
o
f
c
e
n
t
r
a
l
p
i
x
e
l
s
1
4
4
S
i
z
e
o
f
D
N
S
map
2
1
x
2
1
S
i
z
e
o
f
G
N
S
map
2
1
x
2
1
G
a
p
b
e
t
w
e
e
n
t
w
o
c
e
n
t
r
a
l
p
i
x
e
l
s
5
T
o
t
a
l
n
u
m
b
e
r
o
f
d
i
me
n
si
o
n
s
o
f
f
e
a
t
u
r
e
v
e
c
t
o
r
2
1
6
2
.
5
.
T
ra
ini
ng
I
n
t
h
e
tr
ain
in
g
a
n
d
tes
tin
g
,
w
e
u
til
ize
SVM
(
Si
n
g
le
C
la
s
s
)
w
it
h
a
Ga
u
s
s
ian
r
ad
ial
k
er
n
el
f
u
n
ctio
n
[
6
,
7
]
T
h
e
Gau
s
s
ia
n
r
ad
ial
b
asis
k
er
n
el
f
u
n
ctio
n
is
r
ep
r
esen
ted
as:
(
−
)
=
|
−
|
2
2
2
(
5
)
W
h
er
e
(
−
)
is
th
e
k
er
n
el
f
u
n
ctio
n
,
an
d
ar
e
th
e
in
p
u
t
f
ea
t
u
r
e
v
ec
to
r
s
,
an
d
σ
is
a
p
ar
am
eter
s
e
t b
y
u
s
er
s
to
d
eter
m
i
n
e
th
e
e
f
f
ec
ti
v
e
w
id
th
o
f
t
h
e
b
asis
k
er
n
e
l f
u
n
c
tio
n
[
1
4
]
.
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
T
o
v
alid
ate
th
e
p
r
o
p
o
s
ed
m
o
d
el,
C
A
SI
A
-
I
r
is
-
I
n
ter
v
al
d
ataset
V3
.
0
[
1
1
]
h
as
b
ee
n
u
s
ed
.
A
l
s
o
w
e
h
a
v
e
u
s
ed
th
e
s
a
m
e
d
ataset
f
o
r
co
m
p
ar
is
o
n
p
o
r
tio
n
.
T
h
is
d
ata
s
et
is
co
n
s
is
tin
g
o
f
s
i
x
s
u
b
s
et.
Am
o
n
g
th
o
s
e
w
e
h
av
e
co
n
s
id
er
ed
2
0
0
im
ag
e
s
in
to
t
al,
in
w
h
ic
h
w
e
h
a
v
e
u
tili
ze
d
7
0
%
d
ata
f
o
r
tr
ain
in
g
an
d
3
0
%
d
ata
f
o
r
test
in
g
p
u
r
p
o
s
es f
o
r
ea
ch
e
y
e
o
f
th
e
i
n
d
iv
id
u
al.
Fig
u
r
e
5
.
T
h
e
o
u
tp
u
t b
lo
ck
o
f
th
e
w
h
o
le
p
r
o
ce
s
s
to
g
et
t
h
e
G
NS f
o
r
r
ig
h
t e
y
e
Fig
u
r
e
5
d
em
o
n
s
tr
ate
t
h
e
o
u
tp
u
t
b
lo
ck
s
f
o
r
ea
ch
s
tep
s
ac
co
r
d
in
g
to
o
u
r
p
r
o
p
o
s
e
m
o
d
el
f
o
r
th
e
r
ig
h
t
e
y
e
f
r
o
m
d
etec
tio
n
to
GNS.
A
f
ter
t
h
e
d
etec
tio
n
o
f
th
e
e
y
e,
w
e
eli
m
i
n
ate
t
h
e
r
ef
ec
tio
n
an
d
t
h
en
w
e
d
etec
t
t
h
e
i
n
n
er
cir
cle.
Fro
m
th
e
r
u
b
b
er
s
h
ee
t
i
m
a
g
e
o
f
t
h
e
e
y
e,
w
e
g
et
i
m
a
g
in
ar
y
a
n
d
r
ea
l
r
esp
o
n
s
e.
T
h
en
w
e
ca
lc
u
late
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
CE
I
SS
N:
2088
-
8708
A
N
ew A
p
p
r
o
a
ch
o
f I
r
is
Dete
ctio
n
a
n
d
R
ec
o
g
n
itio
n
(
R
u
b
el
B
is
w
a
s
)
2535
GNS
p
ar
t
an
d
f
r
o
m
t
h
at
r
esu
l
t,
w
e
g
et
th
e
r
ea
l,
i
m
a
g
i
n
ar
y
an
d
ab
s
o
lu
te
r
esu
lt
s
.
Fo
r
ir
is
d
etec
tio
n
p
h
ase,
w
e
h
av
e
co
n
s
id
er
ed
th
e
r
ea
l a
n
d
i
m
ag
in
ar
y
v
al
u
es.
T
ab
le
2
.
E
x
p
er
im
e
n
tal
R
e
s
u
l
ts
Deta
ils
f
o
r
T
r
ain
in
g
an
d
T
esti
n
g
f
o
r
B
o
th
E
y
e
f
o
r
T
h
eir
R
ea
l a
n
d
I
m
ag
i
n
ar
y
Val
u
es
Ey
e
C
o
n
si
d
e
r
e
d
V
a
l
u
e
T
r
a
i
n
/
p
e
r
so
n
T
e
st
/
p
e
r
so
n
R
e
c
o
g
n
i
z
e
d
o
u
t
o
f
5
0
p
e
r
so
n
s
P
e
r
c
h
a
n
t
a
g
e
o
f
A
c
c
u
r
e
c
y
L
e
f
t
R
e
a
l
7
0
%
3
0
%
44
8
8
%
L
e
f
t
I
mag
i
n
a
r
y
7
0
%
3
0
%
40
8
0
%
R
i
g
h
t
R
e
a
l
7
0
%
3
0
%
43
8
6
%
R
i
g
h
t
I
mag
i
n
a
r
y
7
0
%
3
0
%
39
7
8
%
C
o
mb
i
n
e
d
R
e
a
l
7
0
%
3
0
%
46
9
2
%
C
o
mb
i
n
e
d
I
mag
i
n
a
r
y
7
0
%
3
0
%
43
8
6
%
A
cc
o
r
d
in
g
to
T
ab
le
2
,
ex
p
er
i
m
en
tal
r
es
u
lt
s
h
o
w
s
th
at
o
u
r
p
r
o
p
o
s
ed
m
o
d
el
g
iv
e
s
9
2
%
a
cc
u
r
ac
y
f
o
r
co
m
b
i
n
ed
r
ea
l
p
o
r
tio
n
r
ec
o
g
n
i
tio
n
a
n
d
8
6
%
ac
cu
r
ac
y
f
o
r
co
m
b
in
ed
i
m
a
g
i
n
ar
y
p
o
r
tio
n
.
Fo
r
ea
ch
ca
s
e,
th
e
r
ea
l
p
ar
t
al
w
a
y
s
ex
h
ib
its
ac
c
u
r
ac
y
o
f
m
o
r
e
th
an
8
6
%
w
h
i
le
th
e
i
m
ag
i
n
ar
y
o
n
e
ex
h
ib
its
al
w
a
y
s
m
o
r
e
t
h
an
7
8
%,
w
h
ic
h
co
n
cl
u
d
es
a
g
o
o
d
a
m
o
u
n
t
o
f
p
er
ce
n
ta
g
e
o
f
ac
c
u
r
ac
y
.
W
e
h
av
e
co
m
p
ar
ed
o
u
r
p
r
o
p
o
s
ed
m
o
d
el
w
it
h
t
w
o
o
th
er
co
n
v
e
n
tio
n
a
l a
p
p
r
o
ac
h
es: o
n
e
is
b
y
L
i M
.
el
a
l.
[
1
7
]
an
d
o
th
er
o
n
e
is
b
y
K.
Sat
h
i
y
ar
aj
a
el
a
l.
[
1
8
]
.
Fig
u
r
e
6
.
T
h
e
co
m
p
ar
is
o
n
ch
a
r
t a
m
o
n
g
o
u
r
m
o
d
el
v
s
.
alg
o
r
it
h
m
1
v
s
.
alg
o
r
i
t
h
m
2
f
o
r
b
o
th
r
ea
l a
n
d
i
m
a
g
in
ar
y
v
alu
e
f
o
r
b
o
th
e
y
e
Fro
m
F
i
g
u
r
e
6
,
w
e
h
a
v
e
s
ee
n
t
h
at
f
o
r
co
m
b
in
ed
r
ea
l
a
n
d
i
m
a
g
in
ar
y
p
ar
t,
o
u
r
p
r
o
p
o
s
ed
m
o
d
el
g
i
v
es
a
b
etter
p
er
f
o
r
m
a
n
ce
t
h
an
t
w
o
o
th
er
s
tated
m
o
d
els
:
L
i
M.
el
al.
[
1
7
]
as
alg
o
r
ith
m
1
a
n
d
K.
Sath
i
y
ar
aj
a
el
al.
[
1
8
]
as
alg
o
r
ith
m
2
.
I
t
is
c
lear
ly
v
is
ib
le
t
h
at
w
h
ile
f
o
r
r
ea
l
p
o
r
tio
n
th
e
al
g
o
r
ith
m
1
a
n
d
2
is
g
iv
in
g
t
h
e
r
es
u
lt
o
f
8
6
%
an
d
8
4
%,
o
u
r
s
o
n
e
is
g
iv
in
g
9
2
%,
w
h
ich
i
s
r
ea
ll
y
s
ig
n
i
f
ica
n
t.
B
esid
e
th
ese,
i
m
ag
in
ar
y
p
o
r
tio
n
i
s
g
i
v
i
n
g
8
6
% r
ec
o
g
n
itio
n
ac
c
u
r
ac
y
w
h
i
le
alg
o
r
ith
m
1
an
d
2
is
g
i
v
i
n
g
8
2
% a
n
d
8
1
%.
4.
CO
NCLU
SI
O
N
I
n
o
u
r
p
r
o
p
o
s
ed
m
o
d
el,
f
o
r
s
eg
m
en
tat
io
n
p
u
r
p
o
s
e
w
e
h
av
e
u
s
ed
Ho
u
g
h
T
r
an
s
f
o
r
m
atio
n
.
T
h
en
Dau
g
m
a
n
’
s
R
u
b
b
er
Sh
h
et
m
o
d
el
is
u
til
ized
f
o
r
n
o
r
m
al
izati
o
n
p
u
r
p
o
s
e.
Fo
r
r
ed
u
cin
g
t
h
e
n
u
m
b
er
o
f
f
ea
tu
r
e
v
ec
to
r
w
e
h
av
e
u
s
ed
GNS
a
n
d
DNS
m
ap
p
i
n
g
,
w
h
ic
h
cr
ea
tes
s
i
g
n
i
f
ica
n
t
c
h
an
g
es
o
n
e
x
p
er
i
m
e
n
tal
r
es
u
lt.
Ov
er
all,
w
e
h
a
v
e
g
o
t 9
2
% a
cc
u
r
ac
y
i
n
r
ea
l p
o
r
tio
n
a
n
d
8
6
%
ac
cu
r
ac
y
i
n
i
m
a
g
i
n
ar
y
p
o
r
tio
n
f
o
r
b
o
th
e
y
e
s
alo
n
g
w
it
h
t
h
e
r
ed
u
ce
d
d
i
m
e
n
s
io
n
o
f
2
1
6
f
ea
tu
r
e
v
ec
to
r
.
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
.
5
,
Octo
b
er
201
7
:
25
30
–
2
5
3
6
2536
RE
F
E
R
E
NC
E
S
[1
]
R.
P
.
W
il
d
e
s "
Iris
Re
c
o
g
n
it
io
n
:
A
n
Em
e
rg
in
g
Bio
m
e
tri
c
Tec
h
n
o
lo
g
y
"
Pro
c
.
IEE
E
v
o
l.
8
5
n
o
.
9
p
p
.
1
3
4
8
-
1
3
6
3
1
9
9
7
.
[2
]
J.
Da
u
g
m
a
n
.
Ho
w
iri
s
re
c
o
g
n
it
io
n
w
o
rk
s
.
Pro
c
e
e
d
in
g
s
o
f
2
0
0
2
In
t
e
rn
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Ima
g
e
Pro
c
e
ss
in
g
,
V
o
l.
1
,
2
0
0
2
.
[3
]
W
il
d
e
s,
R.
P
.
,
A
s
m
u
t
h
,
J.C.
e
t
a
l.
,
"
A
S
y
ste
m
f
o
r
A
u
to
m
a
ted
Iris
Re
c
o
g
n
it
io
n
"
,
Pr
o
c
.
o
f
t
h
e
S
e
c
o
n
d
IEE
E
W
o
rk
sh
o
p
o
n
A
p
p
li
c
a
ti
o
n
s
o
f
C
o
mp
u
ter
Vi
si
o
n
,
1
9
9
4
,
p
p
.
1
2
1
-
1
2
8
.
[4
]
Jia
li
Cu
i,
Yu
n
h
o
n
g
W
a
n
g
,
Ju
n
Zh
o
u
Hu
a
n
g
,
T
ien
iu
T
a
n
a
n
d
Zh
e
n
a
n
S
u
n
,
“
A
n
Iris
I
m
a
g
e
S
y
n
th
e
sis
M
e
th
o
d
Ba
se
d
o
n
P
CA
a
n
d
S
u
p
e
r
-
re
so
lu
ti
o
n
”
,
IEE
E
CS
Pro
c
e
e
d
in
g
s
o
f
th
e
1
7
th
I
n
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Pa
t
t
e
rn
Rec
o
g
n
it
i
o
n
(
ICPR
’0
4
)
.
2
0
0
4
.
[5
]
A
Yu
n
iarti
,
"
Clas
si
f
ica
ti
o
n
a
n
d
n
u
m
b
e
rin
g
o
f
d
e
n
tal
ra
d
io
g
ra
p
h
s
f
o
r
a
n
a
u
to
m
a
ted
h
u
m
a
n
id
e
n
ti
f
ica
ti
o
n
sy
ste
m
,
”
T
EL
KOM
NIKA
T
e
lec
o
mm
u
n
ic
a
ti
o
n
,
C
o
mp
u
ti
n
g
,
El
e
c
tro
n
ics
a
n
d
Co
n
tro
l.
,
v
o
l
.
1
0
,
n
o
.
1
,
p
p
.
1
3
7
-
1
4
6
,
2
0
1
2
.
[6
]
Kh
e
ll
a
h
F
M
.
T
e
x
tu
re
Clas
si
f
ic
a
ti
o
n
Us
in
g
Do
m
in
a
n
t
Ne
ig
h
b
o
rh
o
o
d
S
tru
c
t
u
re
.
IEE
E
T
ra
n
sa
c
ti
o
n
o
n
Ima
g
e
Pro
c
e
ss
in
g
2
0
1
1
;
2
0
(
1
1
)
:3
2
7
0
-
3
2
7
9
.
[7
]
J.
Ud
d
in
,
M
.
Ka
n
g
,
D.
V
.
Ng
u
y
e
n
,
a
n
d
J.
-
M
.
Kim
,
“
R
e
li
a
b
le
fa
u
lt
c
las
si
f
ica
ti
o
n
o
f
in
d
u
c
ti
o
n
m
o
to
rs
u
sin
g
tex
tu
re
f
e
a
tu
re
e
x
tra
c
ti
o
n
a
n
d
a
m
u
lt
icla
ss
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
,
”
M
a
th
e
ma
ti
c
a
l
Pro
b
lem
s
in
En
g
i
n
e
e
rin
g
,
v
o
l
.
2
0
1
4
,
A
rti
c
le ID
8
1
4
5
9
3
,
9
p
a
g
e
s,
2
0
1
4
.
[8
]
L
.
M
a
,
T
.
T
a
n
,
Y.
W
a
n
g
a
n
d
D
.
Zh
a
n
g
,
“
Ef
f
icie
n
t
Iris
Re
c
o
g
n
it
io
n
b
y
Ch
a
ra
c
teriz
in
g
Ke
y
L
o
c
a
l
V
a
riatio
n
s”
,
IE
EE
T
ra
n
sa
c
ti
o
n
s
o
n
Ima
g
e
Pro
c
e
ss
in
g
,
V
o
l
.
1
3
,
N
o
.
6
,
2
0
0
4
,
p
p
.
7
3
9
.
[9
]
IKG
D
P
u
tra,
E
Er
d
iaw
a
n
,
"
Hig
h
p
e
rf
o
rm
a
n
c
e
p
a
l
m
p
rin
t
id
e
n
ti
f
ica
ti
o
n
sy
ste
m
b
a
se
d
o
n
tw
o
d
im
e
n
sio
n
a
l
g
a
b
o
r
,
”
T
EL
KOM
NIKA
T
e
lec
o
mm
u
n
ic
a
ti
o
n
C
o
mp
u
ti
n
g
El
e
c
tro
n
ics
a
n
d
Co
n
tro
l.
,
v
o
l.
8
,
n
o
.
3
,
p
p
.
3
0
9
-
3
1
8
,
2
0
1
0
.
[1
0
]
"
T
h
e
CAS
IA
iris
ima
g
e
d
a
ta
b
a
se
"
,
[
o
n
li
n
e
]
A
v
a
il
a
b
le:
h
tt
p
:/
/b
io
m
e
tri
c
s.id
e
a
lt
e
st.o
rg
.
[1
1
]
D.
J.
F
ield
.
Re
la
ti
o
n
s
b
e
twee
n
th
e
sta
ti
stics
o
f
n
a
t
u
ra
l
ima
g
e
s
a
n
d
th
e
re
sp
o
n
se
p
ro
p
e
rties
o
f
c
o
rtic
a
l
c
e
ll
s
.
J.
Op
t.
S
o
c
.
Am
.
A
,
1
9
8
7
,
p
p
.
2
3
7
9
-
2
3
9
4
.
[1
2
]
Na
se
e
m
I,
A
le
e
m
A
,
T
o
g
n
e
ri
R,
Be
n
n
a
m
o
u
n
M
,
“
Iris
re
c
o
g
n
it
io
n
u
sin
g
c
las
s
-
sp
e
c
i
f
ic
d
ictio
n
a
ries
”
,
Co
mp
u
ter
s
&
El
e
c
trica
l
En
g
in
e
e
rin
g
,
F
e
b
1
8
,
2
0
1
6
.
[1
3
]
A
n
w
a
r,
A
.
M
.
"
An
Iris
d
e
tec
ti
o
n
a
n
d
re
c
o
g
n
i
ti
o
n
sy
ste
m
to
me
a
su
re
th
e
p
e
rfo
rm
a
n
c
e
o
f
E
-
se
c
u
rit
y
.
"
Diss
.
,
BRA
C
Un
iv
e
rsit
y
,
2
0
1
6
.
[1
4
]
G
.
In
d
ra
w
a
n
,
S
.
A
k
b
a
r
a
n
d
B.
S
it
o
h
a
n
g
,
“
F
i
n
g
e
rp
rin
t
Dire
c
t
-
Ac
c
e
ss
S
trate
g
y
U
sin
g
L
o
c
a
l
-
S
tar
-
S
tru
c
tu
re
b
a
se
d
Disc
ri
m
in
a
to
r
F
e
a
tu
re
s:
A
Co
m
p
a
riso
n
S
tu
d
y
,
”
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
m
p
u
ter
En
g
in
e
e
rin
g
(
IJ
ECE
).
”
v
o
l.
4
,
n
o
.
5
,
Oc
to
b
e
r
2
0
1
4
,
p
p
.
8
1
7
-
8
3
0
.
[1
5
]
S
.
S
a
p
a
ru
d
in
,
S
.
A
k
b
a
r
a
n
d
G
.
S
u
lo
n
g
,
“
S
e
g
m
e
n
tatio
n
o
f
F
in
g
e
rp
rin
t
Im
a
g
e
B
a
se
d
o
n
G
ra
d
ien
t
M
a
g
n
it
u
d
e
a
n
d
Co
h
e
re
n
c
e
,
”
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
m
p
u
ter
E
n
g
in
e
e
rin
g
(
IJ
ECE
)”
v
o
l.
5
,
n
o
.
5
,
Oc
to
b
e
r
2
0
1
,
p
p
.
834
-
8
4
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.