I
AE
S
I
nte
rna
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
9
,
No
.
3
,
Sep
tem
b
er
2020
,
p
p
.
429
~
438
I
SS
N:
2252
-
8938
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ai.
v
9
.i
3
.
p
p
429
-
4
3
8
429
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Co
m
pa
riso
n b
etw
een SVM
and
K
N
N clas
sifiers
for ir
is
recog
nition us
ing
a
new
unsuperv
is
ed neural a
pp
ro
a
ch in
seg
m
en
tatio
n
H
icha
m
O
h
m
a
id
1
,
S.
E
dd
a
ro
uich
2
,
A.
B
o
uro
uh
o
u
3
,
M
.
T
im
o
uy
a
s
4
1
Na
ti
o
n
a
l
S
c
h
o
o
l
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
S
y
ste
m
s
A
n
a
l
y
sis (E
NS
IA
S
),
M
o
h
a
m
m
e
d
V
Un
iv
e
rsity
,
Ra
b
a
t/
1
0
1
1
2
,
M
o
ro
c
c
o
2
Re
g
io
n
a
l
Ed
u
c
a
ti
o
n
a
l
Ce
n
ter,
Ra
b
a
t/
1
0
1
1
2
,
M
o
r
o
c
c
o
3,
4
Hig
h
e
r
No
rm
a
l
S
c
h
o
o
l
o
f
T
e
c
h
n
ica
l
Ed
u
c
a
ti
o
n
(ENS
ET
),
M
o
h
a
m
m
e
d
V
Un
iv
e
rsi
ty
,
ENS
ET
,
Ra
b
a
t/
1
0
1
1
2
,
M
o
r
o
c
c
o
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
D
ec
1
2
,
2
0
19
R
ev
i
s
ed
F
eb
17
,
2
0
20
A
cc
ep
ted
M
ar
1
9
,
2
0
20
A
b
io
m
e
tri
c
s
y
ste
m
o
f
id
e
n
ti
f
ica
ti
o
n
a
n
d
a
u
t
h
e
n
ti
c
a
ti
o
n
p
ro
v
id
e
s
a
u
to
m
a
ti
c
re
c
o
g
n
it
io
n
o
f
a
n
in
d
iv
id
u
a
l
b
a
se
d
o
n
c
e
rtain
u
n
i
q
u
e
f
e
a
tu
re
s
o
r
c
h
a
ra
c
teristics
h
e
o
r
sh
e
p
o
ss
e
ss
e
s.
Iris
re
c
o
g
n
it
io
n
is
a
b
io
m
e
t
ric
id
e
n
ti
f
ica
ti
o
n
m
e
th
o
d
th
a
t
a
p
p
li
e
s
p
a
tt
e
rn
re
c
o
g
n
it
io
n
t
o
im
a
g
e
s
o
f
th
e
iri
s.
Ow
in
g
to
th
e
u
n
i
q
u
e
e
p
ig
e
n
e
ti
c
p
a
tt
e
rn
s
o
f
th
e
iri
s,
iri
s
re
c
o
g
n
it
io
n
is
c
o
n
sid
e
re
d
o
n
e
o
f
th
e
m
o
st
a
c
c
u
ra
te
m
e
th
o
d
s
in
th
e
f
ield
o
f
b
i
o
m
e
tri
c
id
e
n
ti
f
ica
ti
o
n
.
T
h
e
se
g
m
e
n
tatio
n
a
lg
o
rit
h
m
p
ro
p
o
se
d
in
th
is
a
rti
c
le
sta
rts
w
it
h
d
e
term
in
in
g
th
e
re
g
io
n
s
o
f
th
e
e
y
e
u
sin
g
u
n
su
p
e
rv
ise
d
n
e
u
ra
l
a
p
p
r
o
a
c
h
,
a
f
ter
th
e
o
u
t
li
n
e
o
f
th
e
e
y
e
is
f
o
u
n
d
u
s
in
g
th
e
Ca
n
n
y
e
d
g
e
,
T
h
e
Ho
u
g
h
T
ra
n
s
f
o
r
m
is
e
m
p
lo
y
e
d
to
d
e
term
in
e
th
e
c
e
n
ter
a
n
d
ra
d
iu
s
o
f
th
e
p
u
p
il
a
n
d
th
e
iri
s.
T
h
e
n
th
e
n
o
rm
a
li
z
a
ti
o
n
a
ll
o
w
s
tran
sf
o
rm
in
g
th
e
se
g
m
e
n
ted
c
ircu
lar
iri
s
re
g
io
n
in
to
a
f
ix
e
d
-
siz
e
re
c
tan
g
u
lar
sh
a
p
e
u
sin
g
Da
u
g
m
a
n
’s
ru
b
b
e
r
sh
e
e
t
m
o
d
e
l.
A
d
isc
re
te
wa
v
e
let
tran
sf
o
r
m
a
ti
o
n
(DWT
)
is
a
p
p
li
e
d
to
th
e
n
o
rm
a
li
z
e
d
iri
s
to
lo
w
e
r
th
e
s
ize
o
f
iri
s
m
o
d
e
ls
a
n
d
to
im
p
ro
v
e
c
las
si
f
ier
a
c
c
u
ra
c
y
.
F
in
a
ll
y
,
th
e
URIBIS
iri
s
d
a
tab
a
se
is
u
se
d
f
o
r
in
d
iv
id
u
a
l
u
se
r
v
e
rif
ica
ti
o
n
b
y
u
sin
g
th
e
KN
N
c
las
si
f
ier
o
r
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(S
VM)
w
h
ich
b
a
se
d
o
n
t
h
e
a
n
a
ly
sis o
f
iri
s c
o
d
e
a
s f
e
a
tu
re
e
x
trac
ti
o
n
is
d
isc
u
ss
e
d
.
K
ey
w
o
r
d
s
:
B
io
m
etr
ic
Dis
cr
ete
w
av
ele
t tr
an
s
f
o
r
m
a
ti
o
n
Ho
u
g
h
tr
an
s
f
o
r
m
I
r
is
s
eg
m
e
n
tatio
n
K
-
n
ea
r
est
-
n
ei
g
h
b
o
r
(
KNN)
Su
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SV
M)
k
er
n
el
f
u
n
ctio
n
Un
s
u
p
er
v
i
s
ed
n
eu
r
al
ap
p
r
o
ac
h
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Hich
a
m
O
h
m
aid
,
Natio
n
al
Sc
h
o
o
l o
f
C
o
m
p
u
ter
Scien
ce
a
n
d
S
y
s
te
m
s
An
al
y
s
i
s
(
E
NSI
A
S),
Mo
h
a
m
m
ed
V
Un
iv
er
s
it
y
,
R
a
b
at/1
0
1
1
2
,
Mo
r
o
cc
o
.
E
m
ail:
h
ic
h
a
m
.
o
h
m
aid
@
g
m
a
i
l
.
co
m
1.
I
NT
RO
D
UCT
I
O
N
No
w
ad
a
y
s
,
b
io
m
etr
ic
r
ec
o
g
n
it
io
n
h
a
s
b
ec
o
m
e
a
r
eliab
le
w
a
y
to
id
en
ti
f
y
a
n
d
r
ec
o
g
n
ize
in
d
iv
id
u
al
s
.
C
o
n
v
en
t
io
n
al
a
u
th
e
n
ticat
io
n
m
et
h
o
d
s
lik
e
p
as
s
w
o
r
d
s
a
n
d
id
en
tific
atio
n
ca
r
d
s
ar
e
n
o
t
al
wa
y
s
cr
ed
ib
le
b
ec
au
s
e
th
ese
m
et
h
o
d
s
ca
n
b
e
f
o
r
g
o
tte
n
o
r
s
to
l
en
.
I
d
en
ti
f
icatio
n
b
as
ed
o
n
p
h
y
s
io
lo
g
ical
o
r
b
eh
av
i
o
r
al
ch
ar
ac
ter
is
tics
o
f
p
eo
p
le
(
s
u
c
h
as
ir
is
es,
f
i
n
g
e
r
p
r
in
ts
,
f
ac
e
s
,
h
an
d
g
eo
m
etr
y
,
s
ig
n
at
u
r
es,
f
i
n
g
er
p
r
in
t
s
,
et
ce
t
er
a)
ca
n
ac
cu
r
atel
y
au
th
e
n
tica
te
p
eo
p
le’
s
id
en
ti
t
y
[
1
]
.
I
r
is
r
ec
o
g
n
itio
n
o
f
f
er
s
t
h
e
h
ig
h
e
s
t
ac
c
u
r
ac
y
i
n
id
e
n
ti
f
y
i
n
g
i
n
d
iv
id
u
als
co
m
p
ar
ed
to
an
y
o
t
h
er
b
io
m
etr
ic
ap
p
r
o
ac
h
es.
T
h
e
ir
is
is
d
ef
i
n
ed
as
t
h
e
cir
cu
lar
p
ar
t
b
etw
ee
n
t
h
e
p
u
p
il
an
d
s
cler
a
o
f
th
e
e
y
e.
I
t
is
s
o
u
n
iq
u
e
th
at
n
o
t
w
o
ir
is
es
ar
e
alik
e,
e
v
en
a
m
o
n
g
id
e
n
t
ical
t
w
in
s
o
r
b
et
w
ee
n
t
h
e
le
f
t
an
d
r
ig
h
t
e
y
e
o
f
t
h
e
s
a
m
e
p
er
s
o
n
[
2
-
3
]
.
T
h
ese
v
is
ib
le
ch
ar
ac
ter
is
tic
s
ar
e
th
o
u
g
h
t
to
b
e
d
is
cr
i
m
in
a
to
r
y
an
d
u
n
iq
u
e
to
ea
c
h
e
y
e,
as
w
e
ll
as
s
tab
le
o
v
er
a
n
in
d
i
v
id
u
al’
s
li
f
eti
m
e.
T
h
is
m
a
k
es
t
h
e
ir
is
a
v
er
y
u
s
e
f
u
l
b
io
m
etr
ic
i
d
en
tifie
r
w
h
e
n
it
is
p
o
s
s
ib
le
to
ca
p
tu
r
e
ir
is
i
m
a
g
es
o
f
s
u
f
f
icie
n
t q
u
alit
y
[
4
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
42
9
–
43
8
430
T
h
e
s
y
s
te
m
i
s
i
n
s
p
ir
ed
b
y
Dau
g
m
a
n
’
s
w
o
r
k
[
5
]
.
I
t
is
c
o
m
p
o
s
ed
o
f
a
n
u
m
b
er
o
f
s
u
b
-
s
y
s
te
m
s
,
w
h
ic
h
ar
e
d
escr
ib
ed
in
Fig
u
r
e
1
an
d
Fig
u
r
e
2
th
e
co
n
tai
n
s
s
ta
g
es o
f
a
n
ir
is
d
etec
tio
n
s
y
s
te
m
.
T
h
e
s
eg
m
en
ta
tio
n
s
tep
en
ab
le
s
th
e
lo
ca
lizatio
n
o
f
t
h
e
ir
is
r
e
g
io
n
i
n
a
n
e
y
e
i
m
ag
e,
w
h
ic
h
c
o
n
s
is
ts
o
f
f
i
n
d
in
g
t
w
o
co
n
to
u
r
s
o
f
th
e
i
r
is
(
p
u
p
il
a
n
d
ir
is
)
.
I
n
o
th
er
w
o
r
d
s
,
t
h
e
i
n
ter
n
a
l
co
n
to
u
r
(
p
u
p
il/i
r
is
)
a
n
d
o
u
ter
(
ir
is
/s
cler
a)
ar
e
lo
ca
ted
.
On
ce
th
e
ir
i
s
r
eg
io
n
is
s
eg
m
e
n
ted
,
th
e
n
ex
t
s
ta
g
e
is
to
n
o
r
m
a
lize
th
i
s
s
e
g
m
en
tat
io
n
.
T
h
e
n
o
r
m
al
izatio
n
tr
a
n
s
f
o
r
m
s
th
e
ir
is
te
x
t
u
r
e
f
r
o
m
C
ar
te
s
ian
to
p
o
lar
co
o
r
d
in
ates
in
o
r
d
er
t
o
lim
it
th
e
p
r
o
b
lem
s
o
f
v
ar
iatio
n
s
i
n
t
h
e
e
y
e
(
li
k
e
t
h
e
o
p
tical
s
ize
o
f
th
e
ir
i
s
o
r
t
h
e
p
o
s
itio
n
o
f
th
e
p
u
p
il
i
n
t
h
e
ir
i
s
)
.
T
h
is
s
tep
is
b
ased
o
n
Dau
g
m
an
'
s
r
u
b
b
er
s
h
ee
t
m
o
d
el
[
5
]
.
Featu
r
e
ex
tr
ac
tio
n
is
a
tech
n
i
q
u
e
u
s
ed
to
ex
tr
ac
t
in
f
o
r
m
atio
n
f
r
o
m
t
h
e
ir
i
s
i
m
a
g
e,
i
n
o
r
d
er
to
lo
w
er
t
h
e
s
ize
o
f
ir
is
m
o
d
els
an
d
i
m
p
r
o
v
e
class
i
f
ier
ac
c
u
r
ac
y
.
T
h
ese
f
ea
t
u
r
es
ca
n
n
o
t
b
e
u
s
ed
to
r
ec
o
n
s
tr
u
ct
i
m
ag
e
s
,
b
u
t
t
h
ese
v
al
u
e
s
ar
e
u
s
e
d
in
clas
s
i
f
icatio
n
.
Fin
all
y
,
m
atc
h
i
n
g
is
p
er
f
o
r
m
ed
b
y
co
m
p
ar
in
g
t
h
e
f
ea
t
u
r
es
o
f
a
te
m
p
late
ir
is
w
it
h
t
h
e
f
ea
t
u
r
e
v
ec
to
r
s
o
f
te
m
p
lates i
n
t
h
e
d
atab
ase,
an
d
a
d
ec
is
io
n
is
f
o
r
m
u
lated
[
6
-
7
].
Fig
u
r
e
1
.
I
r
is
r
ec
o
g
n
itio
n
s
y
s
te
m
Fig
u
r
e
2
.
T
y
p
ical
s
tag
e
s
o
f
ir
is
r
ec
o
g
n
itio
n
T
h
e
m
ai
n
m
o
ti
v
atio
n
in
t
h
i
s
r
esear
ch
is
to
p
r
o
p
o
s
e
a
n
e
w
e
f
f
ec
ti
v
e
an
d
r
o
b
u
s
t
alg
o
r
ith
m
to
s
eg
m
en
t
a
n
d
clas
s
if
icatio
n
th
e
clea
r
o
r
n
o
is
y
ir
is
i
m
ag
e
s
.
T
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
ad
d
ed
a
n
e
w
p
r
e
-
p
r
o
ce
s
s
in
g
s
tep
u
s
i
n
g
a
n
e
w
u
n
s
u
p
er
v
i
s
ed
n
e
u
r
al
ap
p
r
o
a
ch
to
d
iv
id
e
th
e
ir
is
i
m
a
g
e
in
t
w
o
r
eg
io
n
s
n
a
m
el
y
ir
is
a
n
d
e
y
ela
s
h
es
r
e
g
io
n
,
s
cl
er
a
an
d
s
k
i
n
,
i
n
o
r
d
er
to
f
ac
il
itate
t
h
e
d
eter
m
in
at
io
n
o
f
t
h
e
ir
is
co
n
to
u
r
i
n
th
e
p
h
ase
C
an
n
y
ed
g
e
d
etec
t
io
n
[
8
]
.
Un
lik
e
o
th
er
ir
i
s
s
e
g
m
e
n
ta
tio
n
al
g
o
r
ith
m
s
,
th
at
p
r
o
ce
s
s
t
h
e
w
h
o
le
i
m
a
g
e
o
f
th
e
e
y
e
(
w
h
ich
co
n
ta
in
s
th
e
n
o
n
-
ir
i
s
r
eg
io
n
s
th
at
g
en
er
ate
s
eg
m
e
n
tat
io
n
er
r
o
r
s
)
s
u
ch
as
D
au
g
m
a
n
alg
o
r
it
h
m
s
.
T
h
en
,
d
is
cr
ete
w
av
e
let
tr
a
n
s
f
o
r
m
atio
n
(
DW
T
)
is
u
s
ed
f
o
r
e
x
tr
ac
tin
g
t
h
e
o
p
ti
m
u
m
f
ea
t
u
r
es
o
f
ir
is
i
m
a
g
es
an
d
r
ed
u
cin
g
t
h
e
r
u
n
t
i
m
e
o
f
clas
s
if
icatio
n
o
f
t
h
ese
ir
is
te
m
p
lat
es.
Fi
n
all
y
,
a
co
m
p
ar
i
s
o
n
w
as
m
ad
e
b
et
w
ee
n
t
h
e
class
i
f
ier
s
S
VM
an
d
KNN.
2.
M
E
T
H
O
DO
L
O
G
Y
T
h
e
ir
is
r
ec
o
g
n
itio
n
s
y
s
te
m
co
n
s
i
s
ts
o
f
th
e
f
o
llo
w
in
g
s
tep
s
:
I
m
ag
e
ac
q
u
is
itio
n
I
r
is
s
eg
m
e
n
t
atio
n
No
r
m
a
lizatio
n
Featu
r
e
ex
tr
ac
tio
n
C
las
s
i
f
icatio
n
I
n
t
h
e
ap
p
r
o
ac
h
p
r
esen
ted
h
er
e,
s
eg
m
e
n
tatio
n
w
a
s
ac
h
ie
v
ed
u
s
i
n
g
t
h
e
Ho
u
g
h
tr
an
s
f
o
r
m
w
i
th
t
h
e
a
i
m
o
f
lo
ca
tin
g
th
e
ir
is
r
eg
io
n
o
f
an
e
y
e
i
m
ag
e.
T
h
e
n
o
r
m
aliza
t
io
n
en
ab
led
th
e
tr
an
s
f
o
r
m
atio
n
o
f
th
e
s
e
g
m
e
n
t
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
C
o
mp
a
r
is
o
n
b
etw
ee
n
S
V
M a
n
d
K
N
N
cla
s
s
ifie
r
s
fo
r
i
r
is
r
ec
o
g
n
itio
n
u
s
in
g
a
.
.
.
(
Hich
a
m
Oh
ma
id
)
431
cir
cu
lar
ir
is
r
eg
io
n
i
n
to
a
f
i
x
ed
-
s
ize
r
ec
tan
g
u
lar
s
h
ap
e
u
s
i
n
g
Da
u
g
m
an
’
s
r
u
b
b
er
s
h
ee
t
m
o
d
el.
T
w
o
lev
el
d
is
cr
ete
w
a
v
elet
tr
an
s
f
o
r
m
atio
n
(
DW
T
)
w
as
ap
p
lied
to
th
e
n
o
r
m
alize
d
ir
is
f
o
r
f
ea
tu
r
e
ex
tr
ac
t
io
n
.
T
h
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
w
a
s
u
s
ed
to
class
if
y
t
h
e
s
i
m
il
ar
it
y
b
et
w
ee
n
t
h
e
ir
is
te
m
p
lat
es.
Fig
u
r
e
3
s
h
o
w
s
a
d
iag
r
a
m
o
f
t
h
e
ir
is
r
ec
o
g
n
itio
n
s
y
s
te
m
.
Fig
u
r
e
3
.
Diag
r
a
m
o
f
ir
is
r
ec
o
g
n
i
tio
n
s
y
s
te
m
2
.
1
.
I
ris s
eg
m
ent
a
t
io
n
I
n
th
is
s
ec
tio
n
,
w
e
p
r
esen
t
a
n
e
w
ir
is
s
eg
m
e
n
tatio
n
ap
p
r
o
ac
h
u
s
ed
a
s
a
p
r
e
-
p
r
o
ce
s
s
i
n
g
s
tep
,
b
ased
on
n
eu
r
al
co
m
p
etiti
v
e
co
n
ce
p
ts
[
9
]
.
A
s
s
h
o
w
n
i
n
F
i
g
u
r
e
4
,
t
h
is
ap
p
r
o
ac
h
allo
w
s
p
ar
titi
o
n
i
n
g
t
h
e
ir
is
i
m
ag
e
in
t
w
o
r
eg
io
n
s
n
a
m
el
y
ir
i
s
an
d
e
y
ela
s
h
e
s
r
eg
io
n
,
s
cler
a
a
n
d
s
k
i
n
.
T
h
en
,
th
e
o
u
tli
n
e
o
f
t
h
e
e
y
e
is
f
o
u
n
d
u
s
i
n
g
t
h
e
C
an
n
y
ed
g
e,
T
h
e
Ho
u
g
h
T
r
an
s
f
o
r
m
is
e
m
p
lo
y
ed
to
d
eter
m
i
n
e
t
h
e
ce
n
ter
an
d
r
ad
i
u
s
o
f
t
h
e
p
u
p
il
an
d
th
e
ir
is
.
Fig
u
r
e
4
s
h
o
w
s
t
h
e
ir
is
s
eg
m
e
n
tatio
n
s
tep
s
.
Fig
u
r
e
4
.
A
l
g
o
r
ith
m
u
s
ed
f
o
r
i
r
is
r
ec
o
g
n
itio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
42
9
–
43
8
432
2
.
1
.
1
.
I
ris s
eg
m
ent
a
t
io
n us
i
n
g
t
he
un
s
up
er
v
i
s
ed
neura
l a
pp
ro
a
ch
I
n
an
u
n
s
u
p
er
v
i
s
ed
co
n
tex
t,
th
at
m
ea
n
s
w
h
e
n
n
o
p
r
io
r
in
f
o
r
m
atio
n
ab
o
u
t
t
h
e
d
ata
s
a
m
p
le
,
w
e
i
n
v
o
lv
e
t
h
e
n
e
u
r
al
co
m
p
etitiv
e
cl
u
s
ter
in
g
p
r
o
c
ed
u
r
e
to
th
e
I
r
is
s
eg
m
e
n
tatio
n
.
T
h
e
o
b
j
ec
tiv
e
o
f
th
e
s
e
ap
p
r
o
ac
h
es
to
class
if
y
i
n
g
t
h
e
i
m
a
g
e
p
ix
el
s
ac
co
r
d
in
g
to
th
e
ir
d
is
tr
ib
u
tio
n
i
n
th
e
r
ep
r
esen
t
atio
n
s
p
ac
e
an
d
to
ass
i
g
n
t
h
e
m
a
lab
el.
As
ill
u
s
tr
ated
in
Fi
g
u
r
e
5
,
i
t
s
tar
ts
f
ir
s
t
b
y
t
h
e
o
r
g
a
n
izatio
n
I
t
s
tar
ts
f
ir
s
t,
b
y
t
h
e
o
r
g
an
izatio
n
o
f
th
e
p
ix
el
s
o
f
I
r
is
i
m
ag
e
in
an
o
b
s
er
v
atio
n
m
atr
ix
(
ea
ch
r
o
w
r
ep
r
ese
n
t
s
a
p
ix
e
l
a
n
d
ea
ch
co
lu
m
n
r
ep
r
esen
ts
a
n
attr
ib
u
t
e)
to
esti
m
a
te
th
e
u
n
d
er
l
y
i
n
g
p
r
o
b
ab
ilit
y
d
en
s
it
y
f
u
n
ctio
n
(
p
d
f
)
o
f
th
e
p
ix
els
d
is
tr
ib
u
tio
n
u
s
i
n
g
a
n
o
n
-
p
ar
am
etr
ic
esti
m
ato
r
.
I
n
th
e
s
ec
o
n
d
s
tep
,
th
e
p
r
o
ce
d
u
r
e
u
s
es
an
ar
tif
ic
ial
n
e
u
r
al
n
et
w
o
r
k
w
it
h
co
m
p
etit
iv
e
tr
a
in
i
n
g
(
NN
C
T
)
to
ex
tr
ac
t
th
e
lo
ca
l
m
a
x
i
m
a
o
f
t
h
e
p
d
f
.
F
o
llo
w
i
n
g
a
m
o
d
es
d
etec
tio
n
m
eth
o
d
u
s
i
n
g
a
tec
h
n
iq
u
e
to
d
etec
t
th
e
ex
i
s
ti
n
g
in
ter
n
e
u
r
al
co
n
n
ec
t
io
n
[
10
]
.
T
h
e
last
s
tep
is
f
o
r
af
f
ec
tin
g
t
h
e
r
e
m
ai
n
i
n
g
p
i
x
els
to
th
eir
class
e
s
.
Fig
u
r
e
5
.
A
r
ch
itectu
r
e
o
f
t
h
e
s
eg
m
e
n
tat
io
n
p
r
o
ce
d
u
r
e
2
.
1
.
2
.
T
he
esti
m
a
t
io
n o
f
un
d
er
ly
ing
pro
ba
bil
it
y
d
ens
it
y
f
un
ct
io
n
Af
ter
co
n
s
tr
u
cti
n
g
th
e
o
b
s
er
v
a
tio
n
m
atr
i
x
o
f
an
I
r
is
im
a
g
e
p
ix
els
Γ
=
{
1
,
2
,
…
,
}
co
n
s
id
er
in
g
as
a
s
et
o
f
Q
N
-
d
i
m
e
n
s
io
n
a
l
o
b
s
er
v
atio
n
s
(
in
th
is
ca
s
e
N=
3
)
W
ith
X
=
[
,
1
,
,
2
,
…
,
,
,
…
,
,
]
,
q
=1
,
2
,
.
.
.
,
Q
an
d
a
p
r
o
b
ab
ilit
y
d
en
s
it
y
f
u
n
c
tio
n
P
(
X)
,
an
o
n
-
p
ar
a
m
e
tr
ic
m
eth
o
d
b
ased
o
n
T
h
e
P
ar
ze
n
[
11
]
w
in
d
o
w
u
s
i
n
g
to
esti
m
ate
t
h
i
s
u
n
d
er
l
y
in
g
d
e
n
s
it
y
f
u
n
c
tio
n
.
T
h
e
tech
n
iq
u
e
is
a
f
ast
esti
m
atio
n
al
g
o
r
ith
m
t
h
at
i
s
p
r
o
p
o
s
ed
b
y
P
o
s
tair
e
an
d
Vass
eu
r
[
12
]
.
First,
th
e
r
an
g
e
o
f
v
ar
iatio
n
o
f
ea
ch
co
m
p
o
n
e
n
t o
f
t
h
ese
o
b
s
er
v
atio
n
s
i
s
n
o
r
m
al
ized
to
th
e
i
n
ter
v
al
[
0
,
R
]
,
w
h
er
e
R
is
an
i
n
te
g
er
s
u
c
h
as
≥
2
,
b
y
m
ea
n
s
o
f
th
e
tr
a
n
s
f
o
r
m
atio
n
d
ef
i
n
e
d
as:
,
=
(
,
−
,
)
(
,
−
,
)
∗
(
1
)
E
ac
h
ax
i
s
o
f
t
h
e
s
o
n
o
r
m
al
i
ze
d
d
ata
s
p
ac
e
is
th
en
p
ar
tit
io
n
ed
in
to
R
e
x
clu
s
i
v
e
in
ter
v
als
o
f
u
n
i
t
w
id
t
h
.
T
h
is
d
is
cr
etiza
tio
n
d
ef
i
n
es
a
s
et
o
f
h
y
p
er
cu
b
e
o
f
u
n
it
s
i
d
e
le
n
g
t
h
.
E
ac
h
h
y
p
er
cu
b
e
n
o
ted
(
)
,
is
a
s
ite
d
ef
in
ed
b
y
its
N
co
o
r
d
in
ates
1
,
2
,
…
,
,
…
,
w
h
ic
h
ar
e
t
h
e
i
n
t
eg
er
p
ar
ts
o
f
th
e
co
o
r
d
in
ate
s
o
f
its
ce
n
tr
e
X
.
T
o
b
e
m
o
r
e
s
p
ec
if
ic,
let
=
[
1
,
,
2
,
,
…
,
,
,
…
,
,
]
,
q
=1
,
2
,
Q
b
e
th
e
Q
o
b
s
er
v
atio
n
s
i
n
th
e
n
o
r
m
alize
d
s
p
ac
e.
E
ac
h
o
b
s
er
v
atio
n
Yq
i
s
f
o
u
n
d
i
n
s
i
d
e
a
n
o
n
-
e
m
p
t
y
h
y
p
er
cu
b
e
w
it
h
th
e
co
o
r
d
in
ates
=
(
,
)
,
n
=1
,
2
,
N,
w
h
er
e
(
,
)
d
esig
n
ates
th
e
i
n
teg
er
p
ar
ts
o
f
,
.
I
f
s
ev
er
a
l
o
b
s
er
v
atio
n
s
f
all
i
n
th
e
s
a
m
e
h
y
p
er
c
u
b
e,
th
is
o
n
e
ap
p
ea
r
s
m
an
y
t
i
m
e
s
o
n
th
e
lis
t
o
f
n
o
n
–
e
m
p
t
y
h
y
p
er
c
u
b
es.
Fu
r
t
h
er
m
o
r
e,
th
e
n
u
m
b
er
o
f
ti
m
e
s
t
h
e
h
y
p
er
cu
b
e
H(
X)
ap
p
ea
r
s
in
t
h
at
lis
t
i
n
d
icate
s
th
e
n
u
m
b
er
o
f
d
ata
p
o
in
ts
q
[
H(
X)
]
w
h
ic
h
f
al
ls
i
n
th
i
s
h
y
p
er
cu
b
e.
Su
b
s
eq
u
e
n
tl
y
,
t
h
e
v
al
u
e
o
f
t
h
e
lo
ca
l d
en
s
it
y
e
s
ti
m
ated
is
:
(
)
=
[
(
)
]
(
2
)
Sin
ce
t
h
e
v
o
l
u
m
e
o
f
H(
X)
is
e
q
u
al
to
u
n
i
t
y
.
So
,
th
is
f
ast
p
r
o
ce
d
u
r
e
allo
w
s
o
n
l
y
t
h
e
esti
m
at
io
n
o
f
t
h
e
u
n
d
er
l
y
in
g
p
r
o
b
ab
ilit
y
d
en
s
i
t
y
f
u
n
ctio
n
at
th
e
ce
n
ter
s
o
f
t
h
e
n
o
n
-
e
m
p
t
y
h
y
p
er
c
u
b
e
w
h
o
s
e
n
u
m
b
er
n
e
v
er
ex
ce
ed
s
th
e
n
u
m
b
er
Q
o
f
av
a
ilab
le
o
b
s
er
v
atio
n
s
(
p
ix
el
s
)
.
A
t
t
h
e
ce
n
ter
s
o
f
t
h
e
h
y
p
er
cu
b
e
ce
lls
,
w
h
ic
h
ar
e
n
o
t
o
n
th
at
lis
t,
th
e
d
en
s
it
y
e
s
ti
m
ate
s
ar
e
k
n
o
w
n
to
b
e
n
u
ll
.
A
t
t
h
e
en
d
o
f
th
is
f
ast
al
g
o
r
ith
m
,
all
th
e
av
a
ilab
le
in
f
o
r
m
atio
n
f
o
r
clu
s
ter
i
n
g
is
in
t
h
e
d
is
cr
ete
s
et
o
f
esti
m
ated
v
al
u
es o
f
t
h
e
u
n
d
er
l
y
in
g
p
r
o
b
ab
ilit
y
d
en
s
i
t
y
f
u
n
ct
i
o
n
(
)
.
In
i
t
i
a
l
I
r
i
s
i
m
a
g
e
N
o
r
mal
i
z
a
t
i
o
n
D
i
scre
t
i
z
a
t
i
o
n
C
a
l
c
u
l
a
t
i
o
n
o
f
p
d
f
E
st
i
m
a
t
i
o
n
o
f
p
d
f
E
x
t
r
a
c
t
i
o
n
o
f
L
o
c
a
l
M
a
x
i
m
a
S
u
p
e
r
v
i
se
d
c
o
m
p
e
t
i
t
i
v
e
l
e
a
r
n
i
n
g
D
e
t
e
c
t
i
o
n
o
f
m
o
d
e
s
U
n
su
p
e
r
v
i
se
d
c
o
m
p
e
t
i
t
i
v
e
l
e
a
r
n
i
n
g
S
e
g
me
n
t
e
d
I
r
i
s
i
mag
e
C
l
a
sse
s
S
e
g
m
e
n
t
a
t
i
o
n
A
f
f
e
c
t
a
t
i
o
n
o
f
t
h
e
r
e
st
o
f
p
i
x
e
l
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
C
o
mp
a
r
is
o
n
b
etw
ee
n
S
V
M a
n
d
K
N
N
cla
s
s
ifie
r
s
fo
r
i
r
is
r
ec
o
g
n
itio
n
u
s
in
g
a
.
.
.
(
Hich
a
m
Oh
ma
id
)
433
2
.
1
.
3
.
T
he
ex
t
ra
ct
io
n o
f
lo
ca
l
m
a
x
i
m
a
by
neura
l net
w
o
rk
Ass
i
m
ilat
in
g
t
h
e
m
o
d
es
to
th
e
lo
ca
l
m
a
x
i
m
a
o
f
th
e
p
d
f
,
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
u
s
es
th
e
Ne
u
r
al
Net
w
o
r
k
s
w
it
h
C
o
m
p
e
titi
v
e
T
r
ain
i
n
g
(
NNCT
)
[
13
]
.
I
n
th
e
tr
ai
n
in
g
a
lg
o
r
it
h
m
,
w
e
w
o
r
k
o
n
l
y
o
n
t
h
e
p
d
f
b
y
p
r
esen
t
in
g
s
eq
u
e
n
tial
l
y
,
th
e
ce
n
ter
s
o
f
th
e
n
o
n
-
e
m
p
t
y
h
y
p
er
cu
b
e
o
f
t
h
e
s
et
X
to
th
e
n
et
w
o
r
k
,
in
s
tead
o
f
th
e
Q
o
b
s
er
v
atio
n
s.
T
h
e
n
eu
r
al
n
et
w
o
r
k
i
s
co
m
p
o
s
ed
o
f
t
w
o
la
y
er
s
:
t
h
e
i
n
p
u
t
la
y
er
an
d
t
h
e
o
u
tp
u
t
la
y
er
.
T
h
e
f
i
r
s
t
o
n
e
is
m
ad
e
o
f
N
u
n
i
ts
,
n
=1
,
2
,
.
.
.
,
N,
s
u
c
h
th
at
u
n
i
t
is
s
o
licited
b
y
th
e
attr
ib
u
te
X
N
o
f
th
e
n
o
n
-
e
m
p
t
y
h
y
p
er
c
u
b
e
H(
X)
w
h
e
n
th
is
o
n
e
is
p
r
ese
n
ted
to
th
e
n
et
w
o
r
k
.
Ho
w
ev
e
r
,
ea
ch
o
u
tp
u
t
n
eu
r
o
n
m
ater
i
alize
s
a
h
y
p
er
cu
b
e
w
h
ic
h
r
ep
r
esen
t
s
th
e
s
ite
o
f
o
n
e
lo
ca
l
m
ax
i
m
u
m
o
f
t
h
e
p
d
f
in
t
h
e
s
et
X,
an
d
p
r
esen
t
s
it
s
w
ei
g
h
t
b
y
t
h
e
m
ea
n
v
ec
to
r
(
)
,
k
=1
,
2
,
…,
K.
T
h
e
n
u
m
b
er
o
f
th
e
o
u
tp
u
t
n
eu
r
o
n
i
s
f
ir
s
t in
itial
ized
ar
b
itra
r
ily
.
Du
r
in
g
th
e
tr
ain
in
g
p
h
ase,
t
h
e
o
u
tp
u
t
n
e
u
r
o
n
s
e
n
ter
i
n
to
co
m
p
e
titi
o
n
w
it
h
ea
c
h
o
t
h
er
b
y
co
m
p
ar
i
n
g
t
h
e
d
is
tan
ce
[
(
)
,
(
)
]
,
k
=1
,
2
,
…,
K,
b
etw
ee
n
t
h
e
i
n
p
u
t
h
y
p
er
cu
b
e
H(
X)
an
d
ea
c
h
o
u
tp
u
t
n
e
u
r
o
n
(
)
,
th
e
w
i
n
n
er
is
th
e
clo
s
est
o
n
e
t
o
th
e
h
y
p
er
cu
b
e,
t
h
e
n
w
e
co
m
p
ar
e
th
e
v
al
u
e
s
o
f
th
e
p
d
f
as
s
o
ciate
d
to
th
e
w
i
n
n
er
n
eu
r
o
n
(
)
an
d
to
H
(
X)
.
T
h
e
d
is
tan
ce
m
ea
s
u
r
e
u
s
ed
in
t
h
i
s
tr
ai
n
in
g
al
g
o
r
ith
m
is
Ma
h
ala
n
o
b
is
d
is
tan
ce
t
h
at
g
iv
e
s
t
h
e
b
est r
esu
l
ts
f
o
r
th
e
n
o
n
-
Ga
u
s
s
ia
n
d
is
tr
ib
u
tio
n
[
14
]
in
s
tead
o
f
E
u
clid
ia
n
d
is
tan
ce
a
s
in
t
h
e
(
NNCT
)
.
2
.
1
.
4
.
Det
ec
t
io
n o
f
s
ig
nifica
nt
m
o
des
o
f
pd
f
T
h
e
ai
m
o
f
th
i
s
p
h
a
s
e
i
s
to
co
n
n
ec
t
ea
ch
g
r
o
u
p
o
f
th
e
clo
s
es
t
m
o
d
es
i
n
s
u
c
h
a
w
a
y
t
h
at
w
e
g
et
a
m
a
p
w
h
ic
h
p
r
eser
v
es
th
e
s
h
ap
e
an
d
s
tr
u
ctu
r
e
o
f
th
e
clas
s
es
ex
is
tin
g
in
th
e
i
m
a
g
e,
b
y
a
n
i
m
p
r
o
v
ed
C
o
m
p
eti
tiv
e
Heb
b
ian
L
ea
r
n
in
g
m
eth
o
d
(
C
HL
i
m
)
[
10
]
w
h
ic
h
allo
w
s
el
i
m
in
ati
n
g
t
h
e
i
n
f
lu
e
n
ce
o
f
t
h
e
o
u
tp
u
t n
e
u
r
al
n
u
m
b
er
.
T
o
g
en
er
ate
th
e
i
n
d
u
ce
d
Dela
u
n
a
y
tr
ia
n
g
u
latio
n
,
th
e
C
H
L
i
m
,
g
i
v
e
n
th
e
K
m
o
d
es
d
etec
te
d
b
y
C
NN
as
p
r
o
to
ty
p
es
in
R
N,
s
u
cc
es
s
i
v
el
y
ad
d
s
co
n
n
ec
t
io
n
s
a
m
o
n
g
th
e
m
.
T
h
e
m
et
h
o
d
d
o
es
n
o
t
ch
an
g
e
t
h
e
w
ei
g
h
t
o
f
p
r
o
to
ty
p
es,
b
u
t
o
n
l
y
g
e
n
er
ates
to
p
o
lo
g
y
ac
co
r
d
in
g
to
th
ese
p
r
o
to
ty
p
es.
Fo
r
ea
ch
m
o
d
e
H(
)
,
ca
n
b
e
co
n
n
ec
ted
to
t
h
e
t
w
o
clo
s
est
p
r
o
to
ty
p
e
s
b
y
an
ed
g
e
u
s
i
n
g
Ma
h
ala
n
o
b
is
d
is
tan
ce
as
a
m
ea
s
u
r
e
o
f
r
ese
m
b
lan
ce
in
s
tead
o
f
E
u
clid
ian
d
is
ta
n
ce
,
it
w
o
r
k
s
as
an
ac
ti
v
atio
n
f
u
n
c
tio
n
f
o
r
co
m
p
etitio
n
b
et
w
ee
n
n
e
u
r
o
n
s
.
T
h
is
lead
s
t
o
th
e
in
d
u
ce
d
Dela
u
n
a
y
tr
ian
g
u
la
tio
n
,
w
h
ic
h
is
li
m
ited
to
t
h
o
s
e
r
eg
io
n
s
o
f
th
e
in
p
u
t
s
p
ac
e
.
2
.
1
.
5
.
Cla
s
s
if
ica
t
io
n pro
ce
du
re
On
ce
t
h
e
d
if
f
er
e
n
t
m
o
d
e
s
ar
e
id
en
ti
f
ied
,
th
e
c
lass
if
ica
tio
n
m
et
h
o
d
[
15
]
th
at
w
e
u
s
e
i
n
t
h
e
p
r
esen
t
w
o
r
k
co
n
s
is
t
s
f
ir
s
t,
in
d
ef
i
n
i
n
g
t
h
e
p
i
x
els
f
a
l
li
n
g
in
to
an
y
m
o
d
e
o
f
a
co
n
n
ec
ted
s
et
as
t
h
e
p
r
o
to
t
y
p
e
o
f
o
n
e
clu
s
ter
.
T
h
en
,
th
e
r
e
m
ai
n
i
n
g
p
ix
els,
w
h
ic
h
d
o
n
o
t
f
al
l
i
n
o
n
e
o
f
t
h
e
d
etec
ted
m
o
d
es,
a
r
e
ass
i
g
n
ed
to
th
e
clu
s
ter
s
attac
h
ed
to
th
eir
n
e
ar
est
n
eig
h
b
o
r
am
o
n
g
t
h
ese
p
r
o
to
ty
p
es
b
y
m
ea
n
s
o
f
Ma
h
alan
o
b
i
s
d
is
ta
n
ce
.
Fig
u
r
e
6
p
r
esen
ts
t
h
e
p
r
o
ce
d
u
r
e
f
o
r
class
i
f
y
in
g
t
h
e
ir
is
i
m
ag
e
.
Fig
u
r
e
6
.
I
llu
s
tr
atio
n
o
f
th
e
I
r
i
s
s
eg
m
e
n
tatio
n
p
r
o
ce
d
u
r
e
2
.
1
.
6
.
H
o
ug
h t
r
a
ns
f
o
r
m
Ho
u
g
h
tr
an
s
f
o
r
m
is
a
s
ta
n
d
ar
d
im
a
g
e
a
n
al
y
s
i
s
to
o
l
f
o
r
f
i
n
d
in
g
cu
r
v
es
t
h
at
ca
n
b
e
d
ef
in
ed
in
a
p
ar
am
etr
ical
f
o
r
m
s
u
c
h
as
li
n
es,
cir
cles,
p
ar
ab
o
las,
an
d
h
y
p
er
b
o
las
[
16
].
T
h
e
ed
g
e
m
a
p
is
th
en
u
s
ed
i
n
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
42
9
–
43
8
434
v
o
tin
g
p
r
o
ce
s
s
to
m
ax
i
m
ize
t
h
e
d
ef
i
n
ed
Ho
u
g
h
tr
an
s
f
o
r
m
f
o
r
th
e
d
esire
d
co
n
to
u
r
.
C
o
n
s
i
d
er
in
g
t
h
e
o
b
tain
ed
ed
g
e
p
o
in
ts
as (
x
j
;
y
j
)
,
j
=
1
;
2
; ……,
n
,
a
Ho
u
g
h
tr
an
s
f
o
r
m
ca
n
b
e
w
r
itte
n
as:
(
,
,
)
=
∑
ℎ
(
,
,
,
=
1
,
)
(
3
)
W
h
er
e
ℎ
(
,
,
,
,
)
=
{
1
(
,
,
,
,
)
=
0
0
ℎ
(
4
)
T
h
e
li
m
b
u
s
a
n
d
p
u
p
il a
r
e
b
o
th
m
o
d
eled
as c
ir
cle
s
an
d
th
e
p
ar
a
m
etr
ic
f
u
n
ctio
n
g
is
d
e
f
in
ed
a
s
:
(
,
,
,
,
)
=
(
−
)
2
+
(
−
)
2
−
2
(
5
)
Ass
u
m
in
g
a
cir
cle
w
it
h
ce
n
t
er
(
x
c;
y
c)
a
n
d
r
ad
iu
s
r
,
t
h
e
ed
g
e
p
o
in
ts
th
at
ar
e
lo
ca
ted
o
v
er
th
e
cir
cle
r
esu
lt
i
n
a
ze
r
o
v
al
u
e
f
o
r
t
h
e
f
u
n
ctio
n
.
T
h
e
v
al
u
e
o
f
g
is
th
e
n
tr
an
s
f
o
r
m
ed
to
1
b
y
th
e
h
f
u
n
ctio
n
,
w
h
ic
h
r
ep
r
esen
ts
th
e
lo
ca
l
p
a
tter
n
o
f
t
h
e
co
n
to
u
r
.
T
h
e
lo
ca
l
p
atter
n
s
ar
e
th
e
n
u
s
ed
i
n
a
v
o
tin
g
p
r
o
ce
d
u
r
e
u
s
i
n
g
t
h
e
Ho
u
g
h
tr
an
s
f
o
r
m
,
H
,
in
o
r
d
er
to
lo
ca
te
th
e
p
r
o
p
er
p
u
p
il
an
d
li
m
b
u
s
b
o
u
n
d
ar
ies.
I
n
o
r
d
er
to
d
etec
t
th
e
li
m
b
u
s
,
o
n
l
y
v
er
t
ical
ed
g
e
in
f
o
r
m
atio
n
i
s
u
s
ed
.
T
h
e
t
w
o
e
y
elid
s
u
s
u
all
y
co
v
er
t
h
e
u
p
p
er
an
d
lo
w
er
p
ar
ts
;
th
ese
p
ar
ts
co
n
tai
n
h
o
r
izo
n
tal
ed
g
e
in
f
o
r
m
a
tio
n
.
T
h
e
h
o
r
izo
n
tal
ed
g
e
in
f
o
r
m
atio
n
is
u
s
e
d
f
o
r
d
etec
tin
g
th
e
u
p
p
e
r
an
d
lo
w
er
e
y
elid
s
,
w
h
ic
h
ar
e
m
o
d
eled
as p
ar
ab
o
lic
ar
c
s
[
9
]
.
Af
ter
t
h
e
Ho
u
g
h
T
r
an
s
f
o
r
m
p
r
o
ce
s
s
is
co
m
p
lete,
s
ix
p
ar
a
m
eter
s
ar
e
s
to
r
ed
:
t
h
e
r
ad
iu
s
a
n
d
x
a
n
d
y
(
th
e
ce
n
ter
co
o
r
d
in
ates
f
o
r
b
o
th
cir
cles)
.
E
y
elid
s
ar
e
is
o
la
ted
b
y
f
ir
s
t
f
itti
n
g
a
lin
e
to
t
h
e
u
p
p
er
an
d
lo
w
er
e
y
elid
u
s
i
n
g
th
e
li
n
ea
r
Ho
u
g
h
tr
an
s
f
o
r
m
.
A
s
ec
o
n
d
h
o
r
izo
n
t
al
lin
e
is
th
e
n
d
r
a
w
n
th
a
t
i
n
ter
s
ec
ts
w
i
th
th
e
f
ir
s
t
lin
e
at
th
e
ir
i
s
ed
g
e
th
at
i
s
clo
s
est
to
th
e
p
u
p
il.
T
h
is
p
r
o
ce
s
s
is
co
m
p
leted
f
o
r
b
o
th
th
e
to
p
an
d
b
o
tto
m
e
y
elid
s
.
T
h
e
s
ec
o
n
d
h
o
r
iz
o
n
tal
li
n
e
a
llo
w
s
m
ax
i
m
u
m
i
s
o
latio
n
o
f
e
y
elid
r
eg
io
n
s
.
Fi
g
u
r
e
7
s
h
o
w
s
an
e
x
a
m
p
le
o
f
p
r
o
p
er
ly
s
eg
m
e
n
ted
ir
is
.
Fig
u
r
e
7
.
E
x
a
m
p
les o
f
th
e
s
eg
m
en
ted
ir
is
2
.
2
.
I
ris no
r
m
a
liza
t
io
n
T
h
e
s
ize
o
f
th
e
ir
is
m
a
y
c
h
an
g
e
b
ec
au
s
e
o
f
v
ar
iatio
n
i
n
t
h
e
illu
m
i
n
atio
n
,
e
v
e
n
f
o
r
an
ir
is
f
r
o
m
th
e
s
a
m
e
p
er
s
o
n
.
T
h
is
elastic
d
e
f
o
r
m
atio
n
in
ir
is
te
x
t
u
r
e
m
u
s
t
b
e
co
m
p
en
s
a
ted
f
o
r
to
ac
h
iev
e
m
o
r
e
ac
cu
r
ate
r
ec
o
g
n
itio
n
r
es
u
lts
.
No
r
m
aliz
atio
n
is
a
tec
h
n
iq
u
e
to
p
r
ep
ar
e
a
s
eg
m
e
n
ted
ir
is
f
o
r
f
ea
t
u
r
e
ex
tr
ac
t
io
n
a
n
d
tr
an
s
f
o
r
m
s
th
e
s
e
g
m
e
n
te
d
cir
cu
lar
ir
is
r
e
g
io
n
i
n
to
a
f
i
x
ed
-
s
ize
r
ec
ta
n
g
u
lar
s
h
ap
e.
T
h
e
C
ar
tesi
a
n
-
to
-
p
o
lar
tr
an
s
f
o
r
m
o
f
t
h
e
ir
is
r
eg
io
n
is
b
ased
o
n
Dau
g
m
a
n
’
s
r
u
b
b
er
s
h
ee
t
m
o
d
el
[
5
]
,
as illu
s
tr
ated
in
Fi
g
u
r
e
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
C
o
mp
a
r
is
o
n
b
etw
ee
n
S
V
M a
n
d
K
N
N
cla
s
s
ifie
r
s
fo
r
i
r
is
r
ec
o
g
n
itio
n
u
s
in
g
a
.
.
.
(
Hich
a
m
Oh
ma
id
)
435
Fig
u
r
e
8
.
C
ar
tesi
an
-
to
-
p
o
lar
co
o
r
d
in
ate
tr
an
s
f
o
r
m
T
h
e
C
ar
tesi
an
-
t
o
-
p
o
lar
tr
an
s
f
o
r
m
o
f
th
e
ir
i
s
r
eg
io
n
is
m
o
d
ele
d
b
y
(
4
)
.
(
(
,
)
)
,
(
,
)
)
→
(
,
)
(
4
)
w
it
h
(
,
)
=
(
1
−
)
(
)
+
(
)
(
,
)
=
(
1
−
)
(
)
+
(
)
(
5
)
T
h
e
co
o
r
d
in
ates
o
f
t
h
e
in
n
er
an
d
o
u
ter
b
o
u
n
d
ar
ies
al
o
n
g
th
e
θ
d
ir
ec
tio
n
ar
e
x
p
,
y
p
a
n
d
x
i,
y
i,
r
esp
ec
ti
v
el
y
.
I
m
a
g
e
-
n
o
r
m
al
ized
en
h
a
n
ce
m
e
n
t
is
r
e
q
u
ir
ed
to
o
b
tain
ac
cu
r
ate
f
e
atu
r
es
b
ec
au
s
e
th
e
n
o
r
m
alize
d
ir
i
s
p
r
ese
n
ts
m
a
n
y
v
ar
iat
io
n
s
d
u
e
to
v
ar
ia
ti
o
n
s
i
n
lig
h
t
ill
u
m
in
a
tio
n
s
.
B
y
u
s
i
n
g
h
is
to
g
r
a
m
eq
u
aliza
tio
n
,
p
i
x
els
in
ten
s
itie
s
ar
e
s
p
r
ea
d
o
v
er
th
e
e
n
tire
ir
i
s
te
m
p
la
te.
Fi
g
u
r
e
9
s
h
o
w
s
a
n
ir
is
te
m
p
late
w
i
th
h
is
to
g
r
a
m
eq
u
a
lizatio
n
.
(
a)
(
b
)
Fig
u
r
e
9.
(
a)
I
r
is
n
o
r
m
alize
d
i
n
to
p
o
lar
co
o
r
d
in
ates
,
(
b
)
E
n
h
an
ce
m
e
n
t o
f
th
e
ir
i
s
n
o
r
m
al
ize
d
im
a
g
e
2
.
3
.
F
e
a
t
ure
ex
t
ra
ct
io
n
I
n
th
e
e
n
co
d
in
g
s
tag
e,
t
w
o
lev
el
d
is
cr
ete
w
av
ele
t
tr
an
s
f
o
r
m
a
t
io
n
(
DW
T
)
is
ap
p
lied
to
th
e
n
o
r
m
al
ized
ir
is
r
eg
io
n
,
in
o
r
d
er
to
d
eter
m
i
n
e
DW
T
co
ef
f
icie
n
ts
,
b
y
p
ass
in
g
t
h
e
s
ig
n
al
i
n
d
i
f
f
er
e
n
t
f
r
eq
u
e
n
c
y
r
a
n
g
es,
n
a
m
e
l
y
lo
w
-
lo
w
(
L
L
)
,
lo
w
-
h
ig
h
(
L
H)
,
h
i
g
h
-
lo
w
(
HL
)
a
n
d
h
i
g
h
-
h
i
g
h
(
H
H)
,
as
in
d
ic
ated
in
Fi
g
u
r
e
1
0
.
T
h
e
f
r
eq
u
en
c
y
r
a
n
g
e
ca
n
b
e
r
ep
r
ese
n
ted
as L
L
<
L
H
<
H
L
<
HH
[
1
7
-
18
].
T
h
e
L
L
s
u
b
-
b
a
n
d
r
ep
r
esen
ts
t
h
e
f
ea
tu
r
e
c
h
ar
ac
ter
is
tic
s
o
f
t
h
e
ir
is
;
th
is
s
u
b
-
b
an
d
(
L
L
)
i
s
u
s
ed
i
n
t
h
e
class
i
f
icatio
n
p
h
ase.
Fi
g
u
r
e
1
1
in
d
icate
s
th
e
r
eso
l
u
tio
n
o
f
t
h
e
n
o
r
m
a
lized
an
d
en
h
an
ce
d
i
r
is
i
m
a
g
e
(
4
8
x
4
3
2
)
.
Af
ter
ap
p
l
y
i
n
g
DW
T
to
an
e
n
h
an
ce
d
ir
i
s
i
m
a
g
e,
t
h
e
r
eso
l
u
t
io
n
o
f
s
u
b
-
b
an
d
i
s
(
2
4
×
2
1
6
)
.
T
h
is
s
u
b
-
b
an
d
is
u
s
ed
i
n
s
tead
o
f
t
h
e
o
r
i
g
in
a
l
n
o
r
m
alize
d
ir
is
d
ata
to
r
ed
u
ce
t
h
e
r
eso
lu
tio
n
o
f
th
e
ir
is
m
o
d
el,
an
d
th
e
r
u
n
ti
m
e
o
f
th
e
clas
s
i
f
icatio
n
i
s
s
i
m
ila
r
l
y
r
ed
u
ce
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
42
9
–
43
8
436
LL
LH
HL
HH
Fig
u
r
e
1
0
.
I
r
is
tem
p
late
u
s
i
n
g
DW
T
(
a)
(
b
)
Fig
u
r
e
11
.
(
a)
E
n
h
an
ce
m
e
n
t o
f
th
e
ir
is
n
o
r
m
alize
d
i
m
ag
e
,
(
b
)
DW
T
tr
an
s
f
o
r
m
s
n
o
r
m
al
ized
ir
is
i
m
a
g
e
i
n
to
L
L
s
u
b
-
b
a
n
d
s
2
.
4
.
Su
pp
o
rt
v
ec
t
o
r
m
a
chine
SVM
T
h
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
e
(
SVM)
p
er
f
o
r
m
s
p
atter
n
r
e
co
g
n
itio
n
ac
co
r
d
in
g
to
th
e
p
r
in
cip
le
o
f
s
tr
u
ct
u
r
al
r
is
k
m
i
n
i
m
izatio
n
.
T
h
e
d
ev
elo
p
m
e
n
t
o
f
SVM
a
s
a
c
lass
i
f
ier
h
as
t
w
o
m
aj
o
r
asp
ec
ts
:
t
h
e
f
ir
s
t
asp
ec
t
is
to
f
i
n
d
th
e
h
y
p
er
p
lan
e
t
h
at
o
p
tim
a
ll
y
s
ep
ar
ates
th
e
t
w
o
cl
ass
es,
a
n
d
t
h
e
s
ec
o
n
d
asp
ec
t
i
s
t
h
e
tr
an
s
f
o
r
m
atio
n
o
f
th
e
li
n
ea
r
l
y
n
o
n
-
s
ep
ar
ab
le
c
lass
i
f
icatio
n
p
r
o
b
lem
i
n
to
a
lin
ea
r
l
y
s
ep
ar
ab
le
p
r
o
b
lem
[
19
]
.
Giv
e
n
an
in
p
u
t
en
tit
y
v
ec
to
r
lear
n
in
g
s
et
an
d
th
e
class
t
ag
(
x
_
i,
y
_
i)
,
i=1
,
2
…,
l,
w
h
er
e
∈
an
d
y
∈
{1
,
-
1
}.
I
n
th
e
ca
s
e
o
f
a
lin
ea
r
l
y
s
ep
ar
ab
le
p
r
o
b
le
m
,
th
e
s
ep
ar
atin
g
h
y
p
er
p
lan
e
(
w
h
ich
d
e
f
i
n
es
t
h
e
b
o
u
n
d
ar
y
b
et
w
ee
n
cla
s
s
1
an
d
class
2
)
ca
n
b
e
r
ep
r
esen
ted
as f
o
llo
w
s
:
,
,
1
2
+
∑
=
1
(
6
)
Su
b
j
ec
t to
(
(
)
+
)
≥
1
−
;
≥
0
(
7
)
I
n
th
e
ca
s
e
o
f
a
n
o
n
-
li
n
ea
r
l
y
s
ep
ar
ab
le
p
r
o
b
lem
,
th
e
tr
ai
n
i
n
g
v
ec
to
r
s
x
_
i
ar
e
m
ap
p
ed
in
to
h
ig
h
er
d
i
m
e
n
s
io
n
al
s
p
ac
e
b
y
th
e
f
u
n
ctio
n
∅
in
o
r
d
er
to
m
ak
e
t
h
e
p
r
o
b
lem
l
in
ea
r
l
y
s
ep
ar
ab
le.
T
h
e
SVM
th
en
f
in
d
s
a
li
n
ea
r
s
ep
ar
atin
g
h
y
p
er
p
lan
e
w
it
h
t
h
e
m
a
x
i
m
al
m
ar
g
in
i
n
t
h
is
h
i
g
h
er
d
i
m
e
n
s
io
n
al
s
p
ac
e.
T
h
e
p
en
alt
y
p
ar
a
m
eter
o
f
th
e
er
r
o
r
ter
m
i
s
C
>
0
.
Fu
r
t
h
e
r
m
o
r
e,
(
,
)
=
∅
(
)
(
)
is
ca
lled
th
e
k
er
n
el
f
u
n
ctio
n
[
2
0
]
.
L
i
n
ea
r
:
(
,
)
=
(
8
)
P
o
ly
n
o
m
ial:
(
,
)
=
(
+
)
,
>
0
(
9
)
Gau
s
s
ia
n
:
(
,
)
=
e
xp
(
−
‖
−
‖
2
2
2
)
>
0
(
1
0
)
3.
RE
SU
L
T
S
T
h
e
C
las
s
if
icatio
n
L
ea
r
n
e
r
A
p
p
(
i
n
clu
d
ed
i
n
t
h
e
Stati
s
tics
an
d
Ma
c
h
i
n
e
L
ea
r
n
in
g
T
o
o
lb
o
x
f
o
r
M
A
T
L
A
B
)
is
a
n
ap
p
licatio
n
t
h
at
ca
n
b
e
u
s
ed
to
t
r
ain
m
o
d
el
s
to
clas
s
i
f
y
d
ata
u
s
i
n
g
s
u
p
er
v
i
s
ed
m
ac
h
in
e
lear
n
i
n
g
m
et
h
o
d
s
.
T
h
e
SVM
an
d
KNN
c
la
s
s
i
f
ier
s
w
er
e
u
s
ed
f
o
r
o
u
r
ap
p
r
o
ac
h
[
2
0
-
21]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
C
o
mp
a
r
is
o
n
b
etw
ee
n
S
V
M a
n
d
K
N
N
cla
s
s
ifie
r
s
fo
r
i
r
is
r
ec
o
g
n
itio
n
u
s
in
g
a
.
.
.
(
Hich
a
m
Oh
ma
id
)
437
B
y
u
s
i
n
g
C
lass
if
icatio
n
L
e
ar
n
er
,
w
e
ca
n
e
x
p
lo
r
e
d
ata,
s
elec
t
f
ea
t
u
r
es,
s
p
ec
if
y
v
alid
atio
n
s
c
h
e
m
es,
an
d
tr
ain
an
d
e
x
p
o
r
t c
lass
i
f
icat
io
n
m
o
d
els to
th
e
M
A
T
L
A
B
w
o
r
k
s
p
ac
e
to
g
en
er
ate
p
r
ed
ictio
n
s
f
o
r
n
e
w
d
ata.
T
o
ev
alu
ate
th
e
ef
f
ec
ti
v
e
n
es
s
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
te
m
,
a
s
e
t
o
f
e
y
e
i
m
ag
e
s
o
b
tain
ed
f
r
o
m
t
h
e
UB
I
R
I
S
d
atab
ases
w
er
e
u
s
ed
[
22
]
.
Fig
u
r
e
1
2
s
h
o
w
s
an
e
x
a
m
p
le
o
f
t
h
e
e
y
e
i
m
a
g
es c
o
n
tain
ed
i
n
t
h
i
s
d
atab
ase.
T
h
e
to
tal
n
u
m
b
er
o
f
ir
is
c
lass
e
s
i
s
1
0
0
,
an
d
ea
ch
ir
is
cla
s
s
co
n
ta
i
n
s
t
h
r
ee
i
m
ag
e
s
s
to
r
ed
in
t
h
e
d
atab
ase
f
o
r
th
is
e
x
p
er
i
m
e
n
t.
I
n
t
h
is
w
o
r
k
,
3
0
0
i
m
a
g
es
w
er
e
u
s
ed
f
o
r
tr
ain
i
n
g
,
an
d
1
0
0
im
ag
e
s
w
er
e
u
s
ed
f
o
r
test
in
g
p
u
r
p
o
s
es.
F
ig
u
r
e
12
.
E
y
e
i
m
ag
e
s
i
n
th
e
UB
I
R
I
S d
atab
ase
Fo
u
r
k
er
n
el
f
u
n
c
tio
n
s
(
L
i
n
e
ar
,
Qu
ad
r
atic,
C
u
b
ic,
an
d
Me
d
iu
m
Ga
u
s
s
ia
n
)
w
er
e
u
s
ed
,
an
d
th
e
class
i
f
icatio
n
ef
f
icie
n
c
y
o
f
ea
ch
k
er
n
el
w
as
m
ea
s
u
r
ed
.
Sin
c
e
th
e
h
i
g
h
est
clas
s
i
f
icatio
n
ac
cu
r
ac
y
w
as
o
b
tai
n
ed
b
y
a
Me
d
i
u
m
Ga
u
s
s
ia
n
k
er
n
el
,
as
s
h
o
w
n
i
n
T
ab
le
1
,
th
i
s
k
er
n
el
w
a
s
u
s
ed
in
o
u
r
s
y
s
te
m
f
o
r
class
i
f
icat
io
n
a
n
d
r
ec
o
g
n
itio
n
p
u
r
p
o
s
es.
As
th
e
s
i
m
u
latio
n
r
esu
l
t
s
h
o
w
n
in
T
ab
le
2
,
th
e
KNN
alg
o
r
ith
m
ass
u
r
ed
b
etter
th
e
ac
cu
r
ac
y
o
f
9
5
%,
as
w
ell
a
s
th
e
ex
ec
u
tio
n
ti
m
e
is
to
o
lo
w
co
m
p
ar
ed
to
th
at
o
f
th
e
S
VM
alg
o
r
ith
m
.
T
ab
le
1
.
E
f
f
icien
c
y
o
f
v
ar
io
u
s
k
er
n
el
f
u
n
ctio
n
s
(
SVM)
K
e
r
n
e
l
Ty
p
e
C
l
a
ssi
f
i
c
a
t
i
o
n
A
c
c
u
r
a
c
y
(
%)
T
r
a
i
n
i
n
g
T
i
me
(
se
c
)
L
i
n
e
a
r
6
8
.
2
4
3
0
.
8
9
Q
u
a
d
r
a
t
i
c
8
4
.
3
5
3
3
.
0
9
C
u
b
i
c
8
8
.
6
5
8
0
.
8
3
M
e
d
i
u
m G
a
u
ss
i
a
n
8
8
.
9
6
0
2
.
2
8
T
ab
le
2
.
E
f
f
icien
c
y
o
f
KN
N
cl
ass
i
f
ier
C
l
a
ssi
f
i
c
a
t
i
o
n
A
c
c
u
r
a
c
y
(
%)
T
r
a
i
n
i
n
g
T
i
me
(
se
c
)
K
N
N
9
5
.
0
0
1
9
.
9
4
T
h
e
ac
cu
r
ac
y
o
f
o
u
r
p
r
o
p
o
s
ed
tech
n
iq
u
e
is
9
5
%,
is
s
h
o
w
n
i
n
T
ab
le
3
,
w
h
ic
h
is
b
etter
th
a
n
t
h
e
o
t
h
er
m
eth
o
d
s
th
at
ar
e
u
s
ed
f
o
r
r
u
n
ti
m
e
co
m
p
ar
is
o
n
.
T
ab
le
3
.
A
cc
u
r
ac
y
c
o
m
p
ar
is
o
n
o
f
o
u
r
p
r
o
p
o
s
ed
tech
n
iq
u
e
w
i
t
h
o
th
er
s
A
u
t
h
o
r
C
l
a
ssi
f
i
c
a
t
i
o
n
A
c
c
u
r
a
c
y
(
%)
H
O
h
ma
i
d
e
t
a
l
.
(
O
u
r
p
r
o
p
o
se
d
)
95
M
H
H
a
m
d
a
n
d
al
.
[
2
3
]
94
S
G
F
i
r
a
k
e
a
n
d
a
l
.
[
2
4
]
9
2
.
8
5
J p
o
o
n
i
a
a
n
d
a
l
.
[
2
5
]
8
2
.
5
4.
CO
NCLU
SI
O
NS
I
n
th
i
s
p
ap
er
,
s
eg
m
e
n
tatio
n
was
ac
h
ie
v
e
d
u
s
i
n
g
p
r
etr
ea
t
m
e
n
t
s
tep
b
ased
o
n
a
n
u
n
s
u
p
er
v
i
s
ed
n
eu
r
al
ap
p
r
o
ac
h
.
T
h
e
n
o
r
m
aliza
tio
n
e
n
ab
led
th
e
tr
a
n
s
f
o
r
m
atio
n
o
f
t
h
e
s
e
g
m
en
ted
cir
c
u
lar
ir
is
r
e
g
i
o
n
in
to
a
f
i
x
ed
-
s
ize
r
ec
tan
g
u
lar
s
h
ap
e
u
s
in
g
Da
u
g
m
an
’
s
r
u
b
b
er
s
h
ee
t
m
o
d
el.
Dis
cr
ete
w
a
v
elet
tr
a
n
s
f
o
r
m
at
io
n
(
DW
T
)
is
u
s
ed
f
o
r
ex
tr
ac
ti
n
g
th
e
o
p
ti
m
u
m
f
ea
t
u
r
es
o
f
ir
i
s
i
m
a
g
es
a
n
d
r
ed
u
cin
g
th
e
r
u
n
ti
m
e
o
f
cla
s
s
i
f
i
ca
tio
n
o
f
t
h
ese
ir
is
te
m
p
lates.
Fi
n
all
y
,
a
co
m
p
ar
is
o
n
w
as
m
ad
e
b
et
w
ee
n
t
h
e
cla
s
s
i
f
ier
s
SVM
an
d
KNN.
E
x
p
er
i
m
en
tal
e
v
al
u
atio
n
u
s
i
n
g
t
h
e
U
R
I
B
I
S
d
atab
ase
c
lear
l
y
d
e
m
o
n
s
tr
ated
t
h
e
p
r
o
p
o
s
ed
tech
n
iq
u
e
p
er
f
o
r
m
s
in
a
n
e
f
f
icien
t
m
an
n
er
,
esp
ec
iall
y
f
o
r
th
e
KNN
cla
s
s
i
f
ier
th
at
g
a
v
e
an
ac
c
u
r
ac
y
o
f
9
5
%.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
42
9
–
43
8
438
RE
F
E
R
E
NC
E
S
[
1
]
F
a
n
g
,
Bin
E
t
T
a
n
g
,
Yu
a
n
Ya
n
.
I
m
p
ro
v
e
d
c
las
s
sta
ti
stics
e
s
ti
m
a
ti
o
n
f
o
r
s
p
a
rse
d
a
ta
p
r
o
b
lem
s
in
o
ff
li
n
e
sig
n
a
tu
re
v
e
ri
f
ica
ti
o
n
.
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
S
y
ste
ms
,
M
a
n
,
a
n
d
Cy
b
e
rn
e
ti
c
s,
Pa
r
t
C
(
Ap
p
li
c
a
t
io
n
s
a
n
d
Rev
iews
)
.
v
o
l.
3
5
,
n
o
3
,
p
p
.
2
7
6
-
2
8
6
,
2
0
0
5
.
[
2
]
G
a
o
,
X
in
b
o
,
Z
h
o
n
g
,
Ju
a
n
j
u
a
n
,
L
I,
Jie
,
e
t
a
l.
F
a
c
e
sk
e
tch
s
y
n
th
e
sis
a
lg
o
rit
h
m
b
a
se
d
o
n
E
-
HM
M
a
n
d
se
lec
ti
v
e
e
n
se
m
b
le.
IEE
E
T
r
a
n
sa
c
ti
o
n
s
o
n
Circ
u
it
s a
n
d
S
y
ste
ms
fo
r V
i
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
1
8
,
n
o
4
,
p
.
4
8
7
-
4
9
6
,
2
0
0
8
.
[
3
]
Yu
,
L
i,
Zh
a
n
g
,
Da
v
id
,
Et
W
a
n
g
,
Ku
a
n
q
u
a
n
.
T
h
e
re
lativ
e
d
istan
c
e
o
f
k
e
y
p
o
in
t
b
a
se
d
iri
s
re
c
o
g
n
it
io
n
.
P
a
tt
e
rn
R
e
c
o
g
n
it
io
n
,
v
o
l.
4
0
,
n
o
2
,
p
.
4
2
3
-
4
3
0
,
2
0
0
7
.
[
4
]
Zh
a
n
g
,
T
a
ip
in
g
,
F
a
n
g
,
Bin
,
L
iu
,
Wein
in
g
,
e
t
a
l.
T
o
tal
v
a
riatio
n
n
o
rm
-
b
a
se
d
n
o
n
n
e
g
a
ti
v
e
m
a
tri
x
fa
c
to
riza
ti
o
n
f
o
r
id
e
n
ti
f
y
in
g
d
isc
ri
m
in
a
n
t
re
p
re
se
n
t
a
ti
o
n
o
f
im
a
g
e
p
a
tt
e
rn
s.
Ne
u
ro
c
o
mp
u
ti
n
g
,
v
o
l.
7
1
,
n
o
1
0
,
p
.
1
8
2
4
-
1
8
3
1
,
2
0
0
8
.
[
5
]
Da
u
g
m
a
n
,
J.
Bio
m
e
tri
c
p
e
rso
n
a
l
id
e
n
ti
f
ica
ti
o
n
sy
ste
m
b
a
se
d
o
n
iri
s
a
n
a
l
y
sis,
Un
it
e
d
S
tate
s
P
a
ten
t
,
US
5
2
9
1
5
6
0
.
1
9
9
4
.
[
6
]
Da
u
g
m
a
n
,
J.,
“
Ho
w
iris
r
e
c
o
g
n
it
i
o
n
wo
rk
s
,
”
Im
a
g
e
P
ro
c
e
ss
in
g
.
2
0
0
2
.
P
r
o
c
e
e
d
in
g
s.
2
0
0
2
I
n
tern
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
,
v
o
l.
1
,
n
o
.
,
p
p
.
I
-
3
3
,
I
-
3
6
v
o
l.
1
,
2
0
0
2
.
[
7
]
M
a
se
k
,
L
ib
o
r,
e
t
a
l.
Re
c
o
g
n
it
i
o
n
o
f
h
u
m
a
n
iri
s p
a
tt
e
rn
s f
o
r
b
io
m
e
tri
c
id
e
n
ti
f
ica
ti
o
n
.
2
0
0
3
.
[
8
]
Oh
m
a
id
,
Hic
h
a
m
,
Ed
d
a
ro
u
ic
h
,
S
.
,
B
o
u
r
o
u
h
o
u
,
A
.,
e
t
a
l.
“
Iris
s
e
g
m
e
n
tatio
n
u
si
n
g
a
n
e
w
u
n
s
u
p
e
rv
ise
d
n
e
u
ra
l
a
p
p
ro
a
c
h
”
.
I
n
t
J
Art
if
In
tell
(
IJ
AI
)
,
v
o
l.
9
,
n
o
1
,
pp
.
58
-
64
,
2
0
2
0
.
[
9
]
H.
Oh
m
a
id
,
M
.
T
i
m
o
u
y
a
s
a
n
d
S
.
Ed
d
a
r
o
u
ich
,
"
A
c
o
mp
a
r
a
ti
v
e
stu
d
y
o
f
simil
a
rity
me
a
su
re
me
n
t
i
n
a
n
e
w
n
e
u
ra
l
mo
d
e
l
o
f
Un
su
p
e
rv
ise
d
c
lu
ste
rin
g
,"
2
0
1
9
T
h
ird
In
tern
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
In
telli
g
e
n
t
Co
m
p
u
ti
n
g
in
Da
ta
S
c
ien
c
e
s
(ICDS),
M
a
rra
k
e
c
h
,
M
o
r
o
c
c
o
,
2
0
1
9
,
p
p
.
1
-
8
,
d
o
i:
1
0
.
1
1
0
9
/ICD
S
4
7
0
0
4
.
2
0
1
9
.
8
9
4
2
2
4
2
.
[
1
0
]
T
i
m
o
u
y
a
s
M
.
,
E
d
d
a
ro
u
ic
h
S
.
a
n
d
Ha
m
m
o
u
c
h
A
.
“
M
o
d
e
re
g
io
n
d
e
tec
ti
o
n
u
si
n
g
im
p
ro
v
e
d
Co
m
p
e
ti
ti
v
e
He
b
b
ian
L
e
a
rn
in
g
f
o
r
u
n
su
p
e
rv
ise
d
c
lu
ste
rin
g
,
”
En
g
i
n
e
e
rin
g
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
:
A
n
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
(
ES
T
IJ
)
,
IS
S
N:
2
2
5
0
-
3
4
9
8
V
o
l.
7
,
No
.
4
,
p
p
.
2
6
-
35
,
2
0
1
7
.
[
1
1
]
P
a
rz
e
n
,
E
.
“
A
n
Esti
m
a
ti
o
n
o
f
a
P
ro
b
a
b
i
li
ty
De
n
sit
y
F
u
n
c
ti
o
n
a
n
d
M
o
d
e
,
”
A
n
n
.
M
a
th
.
S
t
a
t
.
,
v
o
l.
3
3
,
p
p
.
1
0
6
5
-
1
0
7
6
,
1
9
6
2
.
[
1
2
]
P
o
sta
i
re
,
J.
-
G
.
,
&
V
a
ss
e
u
r,
C.
P
.
A
.
“
A
f
a
st
A
l
g
o
rit
h
m
f
o
r
n
o
n
P
a
r
a
m
e
tri
c
P
ro
b
a
b
il
i
ty
De
n
sit
y
Esti
m
a
ti
o
n
,
”
IEE
E
,
T
ra
n
s.
o
n
Pa
t
ter
n
A
n
a
l.
a
n
d
M
a
c
h
in
e
I
n
tel
.
P
A
M
I
-
4
,
n
6
,
p
p
.
6
6
3
-
666
,
1
9
8
2
.
[
1
3
]
Ed
d
a
ro
u
ich
,
S
.
,
&
S
b
ih
i,
A
.
“
Ne
u
ra
l
Ne
two
rk
fo
r
M
o
d
e
s
De
tec
ti
o
n
in
Pa
tt
e
rn
Cl
a
ss
if
ica
t
io
n
.
”
ICT
IS
’0
7
,
M
o
r
o
c
c
o
,
F
e
z
,
3
-
5
p
p
.
3
0
0
-
3
0
3
,
2
0
0
7
.
[
1
4
]
T
i
m
o
u
y
a
s,
M
.
,
Ed
d
a
ro
u
ic
h
,
S
.
,
a
n
d
Ha
m
m
o
u
c
h
,
A
.
“
A
n
e
w
a
p
p
r
o
a
c
h
o
f
c
l
a
ss
if
ica
ti
o
n
fo
r
n
o
n
-
Ga
u
ss
ia
n
d
istrib
u
ti
o
n
u
p
o
n
c
o
m
p
e
ti
ti
v
e
tra
i
n
i
n
g
,
”
(ICC
S
’1
2
)
,
A
g
a
d
ir,
M
o
ro
c
c
o
,
p
p
.
1
-
6
,
2
0
1
2
.
[
1
5
]
Ed
d
a
ro
u
ich
,
S
.
,
a
n
d
S
b
i
h
i,
A
.
“
Dé
tec
ti
o
n
d
e
s
M
o
d
e
s
p
a
r
A
p
p
ro
c
h
e
Ne
u
ro
n
a
le
p
o
u
r
la
Clas
sif
ica
ti
o
n
d
e
s
D
o
n
n
é
e
s
d
’u
n
M
é
l
a
n
g
e
d
e
s
Distri
b
u
ti
o
n
s
No
rm
a
le
s
,
”
Pro
c
e
e
d
in
g
o
f
t
h
e
6
t
h
Af
ric
a
n
Co
n
fer
e
n
c
e
o
n
Res
e
a
r
c
h
in
C
o
mp
u
ter
S
c
ien
c
e
(
CAR
I’0
2
)
,
Ya
o
u
n
d
é
,
Ca
m
e
ro
o
n
,
p
p
.
6
1
-
6
8
,
2
0
0
2
.
[
1
6
]
W
il
d
e
s,
Rich
a
rd
P
.
Iris
re
c
o
g
n
it
i
o
n
:
a
n
e
m
e
rg
in
g
b
io
m
e
t
ric
tec
h
n
o
lo
g
y
.
Pro
c
e
e
d
in
g
s
o
f
t
h
e
IEE
E
,
v
o
l.
8
5
,
n
o
9
,
p
.
1
3
4
8
-
1
3
6
3
,
1
9
9
7
.
[
1
7
]
M
a
,
L
i,
T
a
n
,
T
ien
iu
,
W
a
n
g
,
Yu
n
h
o
n
g
,
e
t
a
l.
Ef
f
ici
e
n
t
iri
s
re
c
o
g
n
it
io
n
b
y
c
h
a
ra
c
teriz
in
g
k
e
y
lo
c
a
l
v
a
riatio
n
s.
I
EE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Ima
g
e
p
ro
c
e
ss
in
g
,
v
o
l.
1
3
,
n
o
6
,
p
.
7
3
9
-
7
5
0
,
2
0
0
4
.
[
1
8
]
G
a
n
o
rk
a
r,
S
a
n
ja
y
Et
M
e
m
a
n
e
,
M
a
y
u
ri.
“
Iris
re
c
o
g
n
it
i
o
n
u
sin
g
d
is
c
re
te
wa
v
e
let
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
A
d
v
a
n
c
e
s in
En
g
i
n
e
e
rin
g
&
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
4
,
n
o
1
,
p
.
3
5
6
,
2
0
1
2
.
[
1
9
]
Ra
n
a
,
Hu
m
a
y
a
n
Ka
b
ir,
Az
a
m
,
M
d
S
h
a
f
iu
l,
A
k
h
tar
,
M
st
Ra
sh
id
a
,
e
t
a
l.
“
A
f
a
st
ir
is
re
c
o
g
n
it
io
n
s
y
ste
m
th
ro
u
g
h
o
p
ti
m
u
m
f
e
a
tu
re
e
x
tra
c
ti
o
n
”
.
Pee
r
J
Co
mp
u
ter
S
c
ien
c
e
,
v
o
l.
5
,
p
.
e
1
8
4
,
2
0
1
9
.
[
2
0
]
Cz
a
jk
a
,
A
.
,
Bo
wy
e
r,
K.
W
.
,
Kr
u
m
d
ick
,
M
.
,
&
V
i
d
a
lM
a
ta,
R.
G
.
“
Re
c
o
g
n
it
io
n
o
f
im
a
g
e
-
o
rien
tatio
n
-
b
a
se
d
iri
s
sp
o
o
f
in
g
”
.
IEE
E
T
ra
n
sa
c
ti
o
n
s o
n
in
fo
rm
a
t
io
n
F
o
re
n
sic
s a
n
d
S
e
c
u
ri
ty
,
1
2
(
9
),
2
1
8
4
-
2
1
9
6
,
2
0
1
7
.
[
2
1
]
Zh
a
n
g
,
Hu
i
a
n
d
G
u
a
n
,
X
ian
g
f
e
n
g
.
Iris
re
c
o
g
n
it
io
n
b
a
se
d
o
n
g
ro
u
p
i
n
g
KN
N
a
n
d
re
c
tan
g
le
c
o
n
v
e
rsio
n
.
In
:
2
0
1
2
IEE
E
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Co
m
p
u
ter
S
c
ie
n
c
e
a
n
d
A
u
to
m
a
t
io
n
En
g
in
e
e
rin
g
.
IE
EE
.
p
.
1
3
1
-
134
,
2
0
1
2
.
[
2
2
]
“
Ub
iri
s: No
isy
V
isib
le
W
a
v
e
len
g
t
h
Iris
Im
a
g
e
D
a
tab
a
se
s.”
[
On
li
n
e
].
A
v
a
il
a
b
le:
UR
L
:h
tt
p
:/
/i
r
is.d
i
.
u
b
i
.
p
t/
.
[
2
3
]
Ha
m
d
,
M
u
t
h
a
n
a
H.
a
n
d
A
h
m
e
d
,
S
a
m
a
h
K.
Bio
m
e
tri
c
s
y
st
e
m
d
e
sig
n
f
o
r
iri
s rec
o
g
n
it
i
o
n
u
sin
g
i
n
telli
g
e
n
t
a
lg
o
rit
h
m
s.
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
M
o
d
e
rn
Ed
u
c
a
ti
o
n
a
n
d
Co
mp
u
ter
S
c
ien
c
e
,
v
o
l.
1
1
,
n
o
3
,
p
.
9
,
2
0
1
8
.
[
2
4
]
F
irak
e
,
S
a
rit
a
G
.
a
n
d
M
a
h
a
jan
,
P
.
M
.
Co
m
p
a
riso
n
o
f
iri
s
re
c
o
g
n
it
io
n
u
sin
g
g
a
b
o
r
w
a
v
e
let,
p
rin
c
ip
a
l
c
o
m
p
o
n
e
n
t
a
n
a
ly
sis
a
n
d
in
d
e
p
e
n
d
e
n
t
c
o
m
p
o
n
e
n
t
a
n
a
ly
sis.
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
n
o
v
a
ti
v
e
Res
e
a
rc
h
in
Co
mp
u
ter
a
n
d
Co
mm
u
n
ica
ti
o
n
E
n
g
i
n
e
e
rin
g
,
v
o
l.
4
,
n
o
6
,
p
.
1
2
3
3
4
-
1
2
3
4
2
,
2
0
1
6
.
[
2
5
]
J
y
o
ti
,
P
.
,
P
a
rv
a
ti
,
B.
,
S
a
n
d
e
e
p
,
K.
G
.
,
a
n
d
a
l.
Ne
w
i
m
p
ro
v
e
d
fe
a
tu
re
e
x
trac
ti
o
n
a
p
p
ro
a
c
h
o
f
IRIS
re
c
o
g
n
it
io
n
.
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
C
o
mp
u
ter
S
y
ste
ms
,
v
o
l.
3
,
n
o
1
,
p
.
1
-
3
,
2
0
1
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.