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
.
3
,
J
u
n
e
201
9
,
p
p
.
1
7
1
1
~
1
7
1
9
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v9
i
3
.
p
p
1
7
1
1
-
1719
1711
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
M
o
m
ent
inv
a
ria
nt
-
ba
sed
features
for
J
a
w
i chara
cter
rec
o
g
nition
F
it
ri
Arnia
,
K
ha
irun Sa
dd
a
m
i,
K
ha
ir
ul M
un
a
di
De
p
a
rte
m
e
n
t
o
f
El
e
c
tri
c
a
l
a
n
d
Co
m
p
u
ter E
n
g
in
e
e
rin
g
,
S
y
iah
Ku
a
la Un
iv
e
rsit
y
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
2
0
,
2
0
1
8
R
ev
i
s
ed
No
v
2
6
,
2
0
1
8
A
cc
ep
ted
Dec
1
1
,
2
0
1
8
A
n
c
ien
t
m
a
n
u
sc
rip
ts
w
rit
ten
i
n
M
a
la
y
-
A
ra
b
ic
c
h
a
ra
c
ters
,
w
h
ich
a
re
k
n
o
w
n
a
s
"
Ja
w
i
"
c
h
a
ra
c
ters
,
a
r
e
m
o
stl
y
f
o
u
n
d
i
n
M
a
lay
w
o
rld
.
No
w
a
d
a
y
s,
m
a
n
y
o
f
th
e
m
a
n
u
sc
rip
ts
h
a
v
e
b
e
e
n
d
ig
it
a
li
z
e
d
.
Un
li
k
e
Ro
m
a
n
letters
,
th
e
re
is
n
o
o
p
t
ica
l
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
(OCR)
so
f
twa
re
f
o
r
Ja
w
i
c
h
a
ra
c
ters
.
Th
is
a
rti
c
le
p
ro
p
o
se
s
a
n
e
w
a
lg
o
rit
h
m
f
o
r
Ja
w
i
c
h
a
ra
c
t
e
r
re
c
o
g
n
it
io
n
b
a
se
d
o
n
Hu
’s
m
o
m
e
n
t
a
s
a
n
in
v
a
rian
t
fe
a
tu
re
th
a
t
we
c
a
ll
th
e
tree
ro
o
t
(T
R)
a
lg
o
rit
h
m
.
T
h
e
T
R
a
lg
o
rit
h
m
a
ll
o
w
s
e
v
e
r
y
J
a
w
i
c
h
a
ra
c
ter
to
h
a
v
e
a
u
n
iq
u
e
c
o
m
b
in
a
ti
o
n
o
f
m
o
m
e
n
t.
S
e
v
e
n
v
a
lu
e
s
o
f
th
e
Hu
’s
m
o
m
e
n
t
a
re
c
a
lcu
late
d
f
ro
m
a
ll
Ja
wi
c
h
a
ra
c
ters
,
w
h
ich
c
o
n
sist
o
f
3
6
iso
late
d
,
2
7
i
n
it
ial,
2
7
m
id
d
le,
a
n
d
3
5
e
n
d
c
h
a
ra
c
ters
;
th
is
m
a
k
e
s
a
to
tal
o
f
1
2
5
c
h
a
ra
c
ters
.
T
h
e
T
R
a
lg
o
rit
h
m
w
a
s
th
e
n
a
p
p
li
e
d
to
re
c
o
g
n
ize
t
h
e
se
c
h
a
ra
c
ters
.
T
o
a
ss
e
ss
th
e
T
R
a
lg
o
r
it
h
m
,
fi
v
e
c
h
a
ra
c
ters
th
a
t
h
a
d
b
e
e
n
r
o
tate
d
t
o
9
0
o
a
n
d
1
8
0
o
a
n
d
sc
a
led
w
it
h
f
a
c
to
rs
o
f
0
.
5
a
n
d
2
w
e
re
u
se
d
.
Ov
e
r
a
ll
,
th
e
re
c
o
g
n
it
io
n
ra
te
o
f
th
e
T
R
a
lg
o
rit
h
m
w
a
s
9
0
.
4
%
;
1
1
3
o
u
t
o
f
1
2
5
c
h
a
ra
c
ters
h
a
v
e
a
u
n
iq
u
e
c
o
m
b
in
a
ti
o
n
o
f
m
o
m
e
n
t
v
a
lu
e
s,
w
h
il
e
tes
ti
n
g
o
n
ro
tate
d
a
n
d
sc
a
led
c
h
a
ra
c
ters
a
c
h
iev
e
d
8
2
.
1
4
%
re
c
o
g
n
it
io
n
ra
te.
T
h
e
p
ro
p
o
se
d
m
e
th
o
d
sh
o
w
e
d
a
su
p
e
rio
r
p
e
rf
o
r
m
a
n
c
e
c
o
m
p
a
re
d
w
it
h
th
e
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
a
n
d
E
u
c
li
d
ia
n
Dista
n
c
e
a
s c
las
si
fi
e
r.
K
ey
w
o
r
d
s
:
Featu
r
e
ex
tr
ac
tio
n
J
a
w
i c
h
ar
ac
ter
r
ec
o
g
n
itio
n
Mo
m
en
t i
n
v
ar
ian
t
Op
tical
ch
ar
ac
ter
r
ec
o
g
n
itio
n
P
atter
n
class
i
f
ier
T
r
ee
r
o
o
t
(
T
R
)
alg
o
r
ith
m
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
:
Fit
r
i
A
r
n
ia,
Dep
ar
t
m
en
t o
f
E
lectr
ical
an
d
C
o
m
p
u
ter
E
n
g
in
ee
r
i
n
g
,
Facu
lt
y
o
f
E
n
g
i
n
ee
r
i
n
g
,
S
y
ia
h
Ku
ala
U
n
i
v
er
s
it
y
,
J
l.
Sy
ec
h
A
b
d
u
r
r
au
f
No
.
7
,
B
a
n
d
a
A
ce
h
,
I
n
d
o
n
e
s
ia.
E
m
ail:
f
.
ar
n
ia
@
u
n
s
y
iah
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
Han
d
w
r
iti
n
g
c
h
ar
ac
ter
r
ec
o
g
n
itio
n
h
as
b
ec
o
m
e
o
n
e
o
f
t
h
e
m
o
s
t
in
ter
esti
n
g
r
esear
c
h
ar
ea
s
in
th
e
d
ig
ital
i
m
a
g
e
p
r
o
ce
s
s
in
g
f
ield
.
T
h
e
in
cr
ea
s
ed
v
ar
iet
y
o
f
ch
ar
ac
ter
s
to
b
e
r
ec
o
g
n
ized
i
m
p
lie
s
th
at
t
h
e
r
esear
c
h
in
to
ch
ar
ac
ter
r
ec
o
g
n
itio
n
s
ti
ll
g
ain
in
g
h
i
g
h
atte
n
tio
n
an
d
is
s
till
n
ec
es
s
ar
y
.
T
h
e
m
ai
n
s
tr
ea
m
r
esear
ch
i
n
ch
ar
ac
ter
r
ec
o
g
n
i
tio
n
in
cl
u
d
es
L
at
in
,
C
h
in
e
s
e,
A
r
ab
ic,
J
ap
an
ese,
an
d
I
n
d
ian
.
I
n
ad
d
itio
n
,
t
h
e
c
u
r
r
en
t
r
esear
c
h
co
m
p
r
is
e
s
t
h
e
v
ar
iatio
n
o
f
A
r
a
b
ic
ch
ar
ac
ter
s
,
s
u
c
h
as Ur
d
u
,
Far
s
i,
an
d
J
a
w
i.
I
n
d
o
n
esia,
esp
ec
iall
y
A
ce
h
,
h
a
s
m
a
n
y
r
elic
b
o
o
k
s
t
h
at
ar
e
w
r
itten
i
n
J
a
w
i
c
h
ar
ac
ter
s
.
J
a
w
i
o
r
Ma
lay
–
A
r
ab
ic
ter
m
in
o
lo
g
y
r
ef
er
s
to
th
e
lan
g
u
ag
e
w
r
i
tten
i
n
A
r
ab
ic
letter
s
b
u
t
v
er
b
alize
d
i
n
Ma
la
y
lan
g
u
ag
e.
T
h
e
Ma
la
y
lan
g
u
ag
e
h
a
s
b
ee
n
u
s
ed
in
So
u
t
h
ea
s
t
A
s
ia,
i
n
clu
d
in
g
I
n
d
o
n
e
s
ia,
Ma
la
y
s
ia,
B
r
u
n
ei,
Si
n
g
ap
o
r
e,
So
u
t
h
T
h
ailan
d
,
So
u
th
P
h
ilip
p
in
es,
an
d
So
u
t
h
M
y
a
n
m
ar
[
1
]
.
T
h
er
e
is
a
s
lig
h
t
d
i
f
f
er
e
n
ce
b
et
w
ee
n
A
r
ab
ic
an
d
J
a
w
i
ch
ar
ac
ter
,
in
w
h
ic
h
th
e
l
atter
h
as
s
i
x
ad
d
itio
n
al
letter
s
ad
ap
ted
to
th
e
Ma
lay
lan
g
u
a
g
e.
T
h
e
ad
d
itio
n
al
lett
er
s
ar
e
letter
ڤ
f
o
r
“
p
”,
letter
ڽ
f
o
r
“
n
y
”,
letter
ڠ
f
o
r
“
n
g
”,
letter
چ
f
o
r
“
C
a”
,
letter
ڬ
f
o
r
“
Ga”
,
an
d
letter
ۏ
f
o
r
“
Va”
.
Mo
r
eo
v
er
,
J
a
w
i
ch
ar
ac
ter
s
co
m
m
o
n
l
y
d
o
n
o
t
u
s
e
v
o
w
el
m
ar
k
s
(
h
ar
a
k
ah
)
[
2
]
.
T
h
e
lis
t
o
f
J
a
w
i
ch
ar
ac
ter
s
ca
n
b
e
s
ee
n
i
n
T
ab
le
1
,
in
w
h
ic
h
t
h
e
s
i
x
ad
d
iti
o
n
al
ch
ar
ac
ter
s
ar
e
s
h
o
w
n
in
b
o
ld
.
T
h
er
e
ar
e
3
6
is
o
lated
ch
ar
ac
ter
s
,
2
7
in
itial
-
f
o
r
m
a
n
d
m
id
d
le
-
f
o
r
m
ch
ar
ac
te
r
s
,
an
d
3
5
en
d
-
f
o
r
m
c
h
ar
ac
ter
s
.
T
h
e
to
tal
n
u
m
b
er
o
f
ch
ar
ac
ter
s
is
1
2
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
3
,
J
u
n
e
2
0
1
9
:
1
7
1
1
-
1
7
1
9
1712
T
ab
le
1
.
J
aw
i C
h
ar
ac
ter
s
No
C
h
a
r
a
c
t
e
r
N
a
me
I
so
l
a
t
e
d
B
e
g
i
n
n
i
n
g
M
i
d
d
l
e
En
d
1
A
l
i
f
ا
اـ
2
Ba
ب
ـب
ـب
ـ
بـ
3
Ta
ت
ـت
ـت
ـ
تـ
4
T
sa
ث
ـث
ـث
ـ
ثـ
5
Ji
m
ج
ـج
ـ
ج
ـ
جـ
6
Ca
چ
ـچ
ـ
چـ
چـ
7
Ha
ح
ـح
ـ
ح
ـ
حـ
8
K
h
a
خ
ـخ
ـ
خ
ـ
خـ
9
Da
د
دـ
10
D
z
a
ذ
ذـ
11
Ra
ر
رـ
12
Za
ز
زـ
13
S
i
n
س
ـس
ـ
س
ـ
سـ
14
sy
a
ش
ـش
ـ
ش
ـ
شـ
15
S
h
a
ص
ـص
ـ
ص
ـ
صـ
16
D
h
a
ض
ـض
ـ
ض
ـ
ضـ
17
T
h
a
ط
ـط
ـ
طـ
طـ
18
Z
h
a
ظ
ـظ
ـ
ظـ
ظـ
19
A
i
n
ع
ـع
ـع
ـ
عـ
20
G
h
a
i
n
غ
ـغ
ـغ
ـ
غـ
21
Ng
a
ڠ
ـڠ
ـ
ڠـ
ڠـ
22
Fa
ف
ـف
ـ
ف
ـ
فـ
23
Pa
ڤ
ـڤ
ـ
ڤ
ـ
ڤـ
24
Qa
ق
ـق
ـ
ق
ـ
قـ
25
K
a
f
ك
ـک
ـک
ـ
كـ
26
G
a
ڬ
ـڬ
ـ
ڬـ
ڬـ
27
L
a
m
ل
ـل
ـ
لـ
لـ
28
Mim
م
ـم
ـم
ـ
مـ
29
N
u
n
ن
ـن
ـن
ـ
نـ
30
N
y
a
ڽ
ـڽ
ـ
ڽـ
ڽـ
31
W
a
w
و
وـ
32
Va
ۏ
ۏـ
33
H
a
a
ه
ـه
ـه
ـ
هـ
34
Ya
ي
ـي
ـي
ـ
يـ
35
h
a
mz
a
h
ء
ء
36
L
a
mal
i
f
ـ
T
o
t
a
l
36
27
27
35
T
h
e
r
ec
o
g
n
itio
n
o
f
J
a
w
i
c
h
ar
ac
ter
s
is
i
m
p
o
r
tan
t
b
ec
au
s
e
J
a
w
i
ch
ar
ac
ter
s
w
er
e
w
id
el
y
u
s
ed
in
t
h
e
So
u
t
h
ea
s
t
A
s
ian
co
u
n
tr
ies
f
o
r
6
0
0
y
ea
r
s
[
3
]
.
T
h
e
J
a
w
i
ch
ar
a
cter
s
w
er
e
u
s
ed
f
o
r
w
r
iti
n
g
Ma
la
y
lan
g
u
ag
e,
an
d
ca
n
b
e
f
o
u
n
d
o
n
m
a
n
y
to
m
b
s
to
n
es,
cu
r
r
en
c
y
,
r
o
y
al
d
ec
r
ee
s
,
b
o
o
k
s
,
an
d
s
o
o
n
.
Fu
r
th
er
m
o
r
e,
alm
o
s
t
1
5
,
0
0
0
m
an
u
s
cr
ip
ts
ar
e
w
r
itte
n
i
n
J
a
w
i
w
it
h
v
ar
io
u
s
co
n
ten
ts
,
n
a
m
el
y
,
tale
s
,
t
h
eo
lo
g
y
,
a
n
d
p
o
e
m
s
t
h
at
ca
n
b
e
f
o
u
n
d
in
m
u
s
eu
m
s
al
l
o
v
er
th
e
w
o
r
ld
in
,
f
o
r
in
s
tan
ce
,
E
n
g
la
n
d
,
th
e
Neth
er
la
n
d
s
,
Ma
la
y
s
ia,
I
n
d
o
n
esia,
an
d
m
a
n
y
o
th
er
co
u
n
tr
ies.
T
h
e
a
n
cie
n
t
m
an
u
s
cr
ip
ts
n
o
t
o
n
l
y
h
o
ld
a
r
ec
o
r
d
o
f
th
e
h
i
s
to
r
y
a
n
d
cu
ltu
r
e
o
f
So
u
t
h
ea
s
t
Asi
a
n
co
u
n
tr
ies,
b
u
t
also
co
n
tai
n
m
ed
ic
al
p
r
escr
ip
tio
n
s
,
m
ilit
a
r
y
s
tr
ateg
ie
s
,
n
a
v
i
g
atio
n
tec
h
n
iq
u
es,
a
n
d
o
th
er
u
s
e
f
u
l k
n
o
w
led
g
e
[
4
]
.
R
esear
ch
i
n
t
h
e
d
ig
ital
i
m
a
g
e
an
d
co
m
p
u
ter
v
i
s
io
n
f
ield
h
as
y
ield
ed
o
p
tical
ch
ar
ac
ter
r
ec
o
g
n
itio
n
(
OC
R
)
f
o
r
C
h
i
n
ese
[
5
]
,
A
r
ab
i
c
[
6
]
,
[
7
]
,
I
n
d
ian
[
8
]
,
an
d
L
ati
n
c
h
ar
ac
ter
s
[
9
]
,
b
u
t
n
o
O
C
R
s
y
s
te
m
h
a
s
y
et
b
ee
n
p
r
esen
ted
f
o
r
J
a
w
i
c
h
ar
ac
ter
s
.
A
n
OC
R
ap
p
licatio
n
f
r
a
m
e
w
o
r
k
r
eq
u
ir
es
f
o
u
r
s
tep
s
,
n
a
m
el
y
:
(
1
)
p
r
ep
r
o
ce
s
s
in
g
[
1
0
]
,
[
1
1
]
,
[
1
2
]
;
(
2
)
s
eg
m
en
tatio
n
;
(
3
)
f
ea
tu
r
e
ex
tr
ac
tio
n
[
1
3
]
,
[
1
4
]
;
an
d
(
4
)
p
atter
n
class
i
f
fi
ica
tio
n
[
1
5
]
.
A
t
p
r
esen
t,
f
e
w
al
g
o
r
ith
m
s
f
o
r
J
a
w
i
c
h
ar
ac
ter
r
ec
o
g
n
itio
n
ex
i
s
t
an
d
th
ese
r
ec
o
g
n
i
tio
n
alg
o
r
ith
m
s
h
a
v
e
o
n
l
y
b
ee
n
te
s
ted
o
n
a
li
m
ited
n
u
m
b
er
o
f
J
a
w
i
ch
ar
ac
ter
s
.
Fo
r
in
s
tan
ce
,
t
h
e
y
h
av
e
b
ee
n
tes
ted
o
n
l
y
o
n
th
e
i
s
o
lated
ch
ar
ac
ter
s
[
1
6
]
.
I
n
r
ec
en
t
y
ea
r
s
,
ch
ar
ac
ter
r
ec
o
g
n
itio
n
o
f
A
r
ab
ic
v
ar
i
an
ts
,
s
u
c
h
as
Far
s
i
an
d
Ur
d
u
,
h
as
g
ai
n
ed
atten
tio
n
.
Ho
w
e
v
er
,
less
atte
n
tio
n
h
as
b
ee
n
p
aid
t
o
J
a
w
i
ch
ar
ac
ter
s
c
o
m
p
ar
ed
w
ith
Ur
d
u
an
d
Far
s
i.
T
h
u
s
,
t
h
e
d
ev
elo
p
m
en
t
o
f
a
J
a
w
i
ch
ar
ac
ter
r
ec
o
g
n
itio
n
al
g
o
r
ith
m
w
it
h
a
co
m
p
let
e
d
ataset
is
i
n
d
ee
d
r
eq
u
ir
ed
s
o
th
e
an
cie
n
t J
a
w
i
m
an
u
s
cr
ip
ts
ca
n
b
e
u
tili
ze
d
in
e
asier
w
a
y
s
.
T
ab
le
2
s
h
o
w
s
t
h
e
d
etail
s
o
f
th
e
ac
c
u
r
ac
y
lev
el
s
o
f
ch
ar
a
cter
r
ec
o
g
n
itio
n
m
eth
o
d
s
u
s
i
n
g
p
atter
n
class
i
f
ier
s
.
T
h
e
m
eth
o
d
p
r
o
p
o
s
ed
b
y
R
az
a
k
th
at
u
s
ed
Ha
m
m
i
n
g
clas
s
i
f
ica
tio
n
ac
h
ie
v
ed
th
e
h
i
g
h
e
s
t
ac
cu
r
ac
y
lev
els,
w
it
h
9
7
%
r
ec
o
g
n
itio
n
r
ate,
b
u
t
it
w
a
s
n
o
t
r
o
b
u
s
t
to
s
ca
lin
g
o
p
er
atio
n
s
[
1
7
]
.
T
h
e
m
et
h
o
d
s
in
T
ab
le
2
n
ee
d
a
tr
ain
in
g
p
r
o
ce
s
s
.
I
n
co
m
m
o
n
,
th
e
n
u
m
b
er
o
f
tr
ain
i
n
g
c
h
ar
ac
ter
s
is
p
r
o
p
o
r
tio
n
al
to
th
e
r
ec
o
g
n
itio
n
ac
cu
r
ac
y
;
t
h
e
g
r
ea
ter
th
e
n
u
m
b
er
o
f
tr
ain
in
g
c
h
ar
ac
ter
s
,
t
h
e
h
i
g
h
er
th
e
r
ec
o
g
n
it
io
n
ac
c
u
r
ac
y
i
s
.
Mo
r
eo
v
er
,
an
o
th
er
p
r
o
b
lem
o
f
a
p
atter
n
class
i
f
ier
estab
li
s
h
ed
u
s
i
n
g
lo
ca
l
o
p
tim
a
s
o
lu
tio
n
s
,
s
u
c
h
as
n
eu
r
al
n
et
w
o
r
k
s
,
is
th
at
it
m
a
y
d
ec
r
ea
s
e
th
e
ac
c
u
r
ac
y
lev
e
l
[
1
8
]
.
O
n
th
e
o
th
er
h
an
d
,
m
et
h
o
d
s
s
u
ch
as
Su
p
p
o
r
t
Vec
to
r
Ma
ch
i
n
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
Mo
men
t in
va
r
ia
n
t
-
b
a
s
ed
fe
a
tu
r
es fo
r
J
a
w
i c
h
a
r
a
cter reco
g
n
itio
n
(
F
itr
i A
r
n
ia
)
1713
(
SVM)
an
d
Hid
d
en
Ma
r
k
o
v
Mo
d
els
(
HM
M)
th
at
ar
e
b
ase
d
o
n
g
lo
b
al
o
p
tim
u
m
s
o
l
u
tio
n
s
,
r
eq
u
ir
e
lo
n
g
er
ti
m
e
an
d
m
o
r
e
m
e
m
o
r
y
to
ex
ec
u
te
th
e
p
r
o
g
r
a
m
.
T
h
e
A
d
ab
o
o
s
t
alg
o
r
ith
m
f
ails
d
u
e
to
n
o
i
s
e.
I
n
A
d
ab
o
o
s
t,
th
er
e
ar
e
t
w
o
p
o
in
ts
o
f
w
ea
k
n
ess
:
th
e
lear
n
in
g
p
r
o
ce
s
s
an
d
class
i
f
ic
atio
n
p
r
o
ce
s
s
[
1
9
]
.
A
f
ter
all,
th
e
r
esu
lt
o
f
ea
ch
p
atter
n
clas
s
i
f
ier
is
i
n
f
l
u
e
n
ce
d
b
y
t
h
e
f
ea
t
u
r
e
s
elec
t
io
n
[
2
0
]
.
T
h
e
m
et
h
o
d
p
r
o
p
o
s
ed
b
y
Her
y
an
to
et
al.
w
as a
b
l
e
to
r
ea
ch
9
5
.
6
7
% r
ec
o
g
n
itio
n
r
ate,
b
u
t th
e
tes
tin
g
p
r
o
ce
s
s
r
ea
ch
ed
an
ac
cu
r
ac
y
o
f
7
3
.
5
9
% [
2
1
]
.
T
ab
le
2
.
P
er
f
o
r
m
a
n
ce
o
f
P
atter
n
C
las
s
i
f
ier
in
J
a
w
i
C
h
ar
ac
ter
R
ec
o
g
n
itio
n
No
R
e
se
a
r
c
h
M
e
t
h
o
d
s
S
c
o
p
e
a
p
p
l
i
c
a
t
i
o
n
R
e
c
o
g
n
i
t
i
o
n
A
c
c
u
r
a
c
y
1
N
a
sr
u
d
i
n
[
1
4
]
I
n
v
a
r
i
a
n
t
F
u
t
u
r
e
,
T
r
a
c
e
t
r
a
n
sf
o
r
m,
M
L
P
I
so
l
a
t
e
d
p
r
i
n
t
e
d
Jaw
i
N
o
t
d
e
t
a
i
l
s
2
N
a
sr
u
d
i
n
[
1
6
]
T
r
i
p
l
e
F
e
a
t
u
r
e
,
S
V
M
,
M
L
P
,
B
a
y
e
si
a
n
,
e
t
c
H
a
n
d
w
r
i
t
t
e
n
J
a
w
i
7
2
.
0
2
%
3
R
a
z
a
k
[
1
7
]
D
i
scre
t
e
W
a
v
e
l
e
t
T
r
a
n
sf
o
r
m (D
W
T
)
,
H
a
mm
i
n
g
c
l
a
ss
i
f
i
c
a
t
i
o
n
Jaw
i
9
7
%
4
H
e
r
y
a
n
t
o
[
2
1
]
F
e
a
t
u
r
e
V
e
c
t
o
r
,
H
y
b
r
i
d
A
N
N
,
D
y
n
a
mi
c
p
r
o
g
r
a
mm
i
n
g
Jaw
i
w
o
r
d
i
so
l
a
t
e
d
T
r
a
i
n
i
n
g
=
9
5
.
6
7
%
T
e
st
i
n
g
=
7
3
.
5
9
%
5
R
e
d
i
k
a
[
2
2
]
F
e
a
t
u
r
e
V
e
c
t
o
r
,
H
M
M
Jaw
i
w
o
r
d
s
8
4
%
6
N
a
sr
u
d
i
n
[
2
3
]
T
r
i
p
l
e
F
e
a
t
u
r
e
TT
,
EC
,
C
C
I
so
l
a
t
e
d
P
r
i
n
t
e
d
a
n
d
h
a
n
d
w
r
i
t
t
e
n
Jaw
i
P
r
i
n
t
e
d
=
8
0
.
5
6
%
H
a
n
d
w
r
i
t
t
e
n
=
6
9
,
4
4
%
Ma
ch
i
n
e
lear
n
in
g
al
g
o
r
ith
m
s
s
u
ch
as
A
r
t
if
ic
ial
Ne
u
r
al
Net
wo
r
k
(
A
N
N)
an
d
SVM
h
av
e
b
ee
n
u
s
ed
t
o
class
i
f
y
a
p
atter
n
,
an
d
w
id
el
y
ap
p
lied
in
ch
ar
ac
ter
r
ec
o
g
n
iti
o
n
[
2
4
]
.
I
t
is
u
s
ed
to
p
er
f
o
r
m
th
e
cla
s
s
i
f
icatio
n
o
r
r
ec
o
g
n
itio
n
i
n
a
p
h
a
s
e
k
n
o
w
n
as
t
h
e
tr
ai
n
in
g
p
h
ase.
T
h
e
tr
ain
i
n
g
m
e
th
o
d
ca
te
g
o
r
izes
t
h
e
ex
tr
ac
ted
f
ea
tu
r
e
s
in
to
s
ev
er
al
cla
s
s
e
s
o
f
s
i
m
ilar
p
atter
n
s
.
T
h
e
A
NN
h
a
s
b
ee
n
u
s
ed
s
in
ce
t
h
e
1
9
8
0
s
,
an
d
at
th
e
en
d
o
f
1
9
9
0
s
SVM
w
a
s
in
tr
o
d
u
ce
d
as
n
e
w
an
d
b
etter
class
i
f
icatio
n
m
et
h
o
d
[
2
5
]
.
T
h
e
d
r
a
w
b
ac
k
o
f
m
a
ch
in
e
lear
n
in
g
is
th
e
m
et
h
o
d
r
eq
u
ir
es
e
x
p
er
i
m
e
n
tat
io
n
m
eth
o
d
to
m
o
d
el
t
h
e
tr
ai
n
in
g
p
h
a
s
e
[
2
6
]
.
SVM
as
p
ar
t
o
f
m
ac
h
in
e
lear
n
i
n
g
m
et
h
o
d
s
h
o
w
s
m
o
r
e
s
u
p
er
io
r
p
er
f
o
r
m
a
n
ce
in
c
h
ar
ac
ter
r
ec
o
g
n
i
tio
n
s
y
s
te
m
th
a
n
o
th
er
m
et
h
o
d
s
s
u
c
h
as
ANN
an
d
K
-
n
ea
r
est
Neig
h
b
o
r
(
KNN)
[
2
7
]
.
Ho
w
ev
er
,
th
e
S
VM
p
r
o
ce
s
s
es
in
f
o
r
m
atio
n
as
a
h
o
li
s
tic
p
atter
n
an
d
f
ail
s
in
tr
ai
n
i
n
g
lo
ca
l
p
atter
n
.
F
u
r
t
h
er
m
o
r
e,
SVM
co
n
s
u
m
e
s
m
o
r
e
s
to
r
ag
e
a
n
d
r
eq
u
ir
es
m
o
r
e
co
m
p
lex
co
m
p
u
ta
tio
n
th
an
o
t
h
er
lear
n
i
n
g
alg
o
r
it
h
m
s
s
u
c
h
as
A
NN
a
n
d
HM
M.
An
o
th
er
d
is
ad
v
an
tag
e
o
f
SVM
i
s
th
e
r
eq
u
ir
e
m
e
n
t
to
d
eter
m
in
e
t
h
e
k
er
n
el
m
eth
o
d
s
.
I
n
t
h
is
ar
ticle,
w
e
p
r
o
p
o
s
e
a
n
o
v
el
tec
h
n
iq
u
e
f
o
r
J
a
w
i
ch
ar
ac
ter
r
ec
o
g
n
itio
n
s
y
s
te
m
,
an
d
w
e
r
ef
er
it
to
as
th
e
T
r
ee
R
o
o
t
(
T
R
)
A
lg
o
r
ith
m
.
T
h
e
alg
o
r
it
h
m
d
o
es
n
o
t
u
s
e
co
m
p
lex
m
at
h
e
m
a
tica
l
co
m
p
u
tatio
n
,
b
u
t
p
er
f
o
r
m
s
s
i
m
p
le
class
if
icatio
n
p
r
o
ce
s
s
an
d
r
eq
u
ir
es
less
s
to
r
ag
e
m
e
m
o
r
y
.
I
t
d
o
es
n
o
t
r
eq
u
ir
e
tr
ain
in
g
p
r
o
ce
s
s
;
it
s
av
es
f
ea
t
u
r
e
te
m
p
late
f
o
r
r
ec
o
g
n
izi
n
g
p
r
o
ce
s
s
.
T
h
e
m
et
h
o
d
w
a
s
test
ed
w
ith
a
d
ataset
o
f
J
a
w
i
h
an
d
w
r
iti
n
g
w
it
h
t
h
e
co
m
p
l
ete
c
h
ar
ac
ter
s
et:
3
6
is
o
lated
,
2
7
in
itial
-
f
o
r
m
,
2
7
en
d
-
f
o
r
m
,
a
n
d
3
5
m
id
d
le
-
f
o
r
m
ch
ar
ac
ter
s
.
T
h
e
f
o
llo
w
i
n
g
s
ec
tio
n
d
is
c
u
s
s
es
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
.
I
n
s
e
ctio
n
t
h
r
ee
,
w
e
d
is
c
u
s
s
th
e
ex
p
er
im
e
n
tal
p
r
o
ce
s
s
o
f
t
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
.
S
ec
tio
n
f
o
u
r
p
r
ese
n
ts
an
d
d
i
s
c
u
s
s
es
t
h
e
r
e
s
u
l
ts
,
a
n
d
f
i
n
all
y
,
th
e
co
n
cl
u
s
io
n
i
s
p
r
esen
ted
in
s
ec
t
io
n
f
iv
e.
2.
P
RO
P
O
SE
D
M
E
T
H
O
D
I
n
d
ig
ital
i
m
ag
e
p
r
o
ce
s
s
i
n
g
an
d
co
m
p
u
ter
v
i
s
io
n
,
t
h
e
m
o
m
e
n
t
in
v
ar
ian
t
is
m
o
s
tl
y
u
s
ed
f
o
r
d
escr
ib
in
g
o
b
j
ec
t
ch
ar
ac
ter
is
tics
.
T
h
er
e
a
r
e
t
w
o
t
y
p
es
o
f
f
ea
tu
r
e
e
x
tr
ac
t
io
n
s
b
ased
o
n
m
o
m
e
n
t
i
n
v
ar
ia
n
ts
:
co
n
to
u
r
-
b
ased
o
r
s
h
ap
e
-
b
ased
m
o
m
en
t in
v
ar
i
an
ts
[
2
8
]
an
d
r
eg
io
n
-
b
ased
m
o
m
en
t in
v
ar
ia
n
ts
[
2
9
]
.
Mo
m
e
n
t
in
v
ar
ia
n
t
s
ar
e
v
er
y
u
s
e
f
u
l
f
o
r
ex
tr
ac
ti
n
g
o
b
j
ec
t
f
ea
tu
r
es
w
it
h
u
n
iq
u
e
c
h
ar
ac
ter
is
tics
r
eg
ar
d
less
o
f
lo
ca
tio
n
,
s
iz
e,
an
d
o
r
ien
tatio
n
.
Hu
p
r
o
p
o
s
ed
s
ev
en
in
v
ar
ian
t
m
o
m
e
n
ts
,
k
n
o
w
n
a
s
Geo
m
et
r
ic
Mo
m
e
n
t
I
n
v
ar
ian
ts
(
GM
I
)
th
at
ar
e
in
v
ar
ia
n
t
to
w
ar
d
tr
an
s
la
tio
n
,
r
o
tatio
n
,
s
c
alin
g
,
an
d
m
ir
r
o
r
in
g
a
s
f
o
llo
ws [
3
0
]
.
ϕ
1
=
η
20
+
η
02
ϕ
2
=
(
η
20
+
η
02
)
2
+
4
η
11
2
ϕ
3
=
(
η
30
−
3
η
12
)
2
+
(
3
η
21
−
η
03
)
2
ϕ
4
=
(
η
30
+
η
12
)
2
+
(
η
21
+
η
03
)
2
ϕ
5
=
(
η
30
−
3
η
12
)
(
η
30
+
η
12
)
[
(
η
30
+
η
12
)
2
−
3
(
η
21
+
η
03
)
2
]
+
(
3
η
21
−
η
03
)
(
η
21
+
η
03
)
[
3
(
η
30
+
η
12
)
2
−
(
η
21
+
η
03
)
2
]
ϕ
6
=
(
η
20
−
η
02
)
[
(
η
30
+
3
η
12
)
2
−
3
(
η
21
−
η
03
)
2
]
+
4
η
11
(
η
30
+
η
12
)
(
η
21
+
η
03
)
ϕ
7
=
(
3
η
21
−
η
03
)
(
η
30
+
η
12
)
[
(
η
30
+
η
12
)
2
−
3
(
η
21
+
η
03
)
2
]
+
(
3
η
12
−
η
30
)
(
η
21
+
η
03
)
[
3
(
η
30
+
η
12
)
2
−
(
η
21
+
η
03
)
2
]
(
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
3
,
J
u
n
e
2
0
1
9
:
1
7
1
1
-
1
7
1
9
1714
2
.
1
.
T
he
t
re
e
ro
o
t
(
T
R)
a
lg
o
ri
t
hm
B
ased
o
n
m
o
m
en
t
v
alu
e
s
(
1
)
,
w
e
d
ev
elo
p
ed
th
e
tr
ee
r
o
o
t
(
T
R
)
alg
o
r
ith
m
.
I
n
t
h
e
alg
o
r
ith
m
,
d
en
o
tes
th
e
ℎ
m
o
m
e
n
t,
w
ith
=
{
1
,
2
,
3
,
4
,
5
,
6
,
7
}
an
d
r
ef
er
s
to
a
g
r
o
u
p
o
f
ch
ar
ac
ter
s
w
it
h
th
e
ℎ
m
o
m
e
n
t a
n
d
m
o
m
e
n
t
v
al
u
e
.
Fo
r
ex
a
m
p
le,
th
e
c
h
ar
ac
ter
g
r
o
u
p
o
f
f
ir
s
t
m
o
m
e
n
t
w
it
h
=
1
is
d
en
o
ted
as
11
.
T
h
e
T
R
alg
o
r
ith
m
is
e
x
p
lain
ed
as f
o
llo
w
s
.
1.
C
alcu
late
s
e
v
en
m
o
m
e
n
t
v
al
u
e
s
o
f
all
ch
ar
ac
ter
s
.
2.
R
o
u
n
d
th
e
m
o
m
e
n
t v
a
lu
e
to
t
h
e
s
m
alle
s
t
in
teg
er
n
u
m
b
er
.
3.
Set
=
1
.
4.
Gr
o
u
p
all
ch
ar
ac
ter
s
b
ased
o
n
th
eir
m
o
m
e
n
t v
al
u
es
n
a
m
el
y
.
5.
C
o
u
n
t
t
h
e
m
e
m
b
er
s
o
f
all
,
if
is
u
n
iq
u
e,
th
e
n
s
av
e
th
e
m
o
m
en
t
v
alu
e
an
d
an
d
t
h
e
c
h
ar
ac
ter
is
u
n
iq
u
e
an
d
r
ec
o
g
n
ized
.
O
th
er
w
i
s
e,
=
+
1
.
6.
C
h
ec
k
v
alu
e,
i
f
=
7
,
th
e
ch
ar
ac
ter
in
th
is
g
r
o
u
p
is
u
n
r
ec
o
g
n
izab
l
e.
Oth
er
w
is
e,
r
etu
r
n
to
s
tep
4
.
3.
RE
S
E
ARCH
M
E
T
H
O
D
I
n
t
h
is
s
ec
tio
n
,
w
e
p
r
esen
t
t
h
e
r
esear
ch
m
et
h
o
d
f
o
r
test
i
n
g
th
e
T
R
alg
o
r
it
h
m
.
T
h
e
e
x
p
er
i
m
e
n
t
w
a
s
co
n
d
u
cted
in
th
r
ee
s
ta
g
es:
(
1
)
d
ata
co
llectio
n
an
d
tem
p
late
p
r
o
d
u
ctio
n
;
(
2
)
f
ea
tu
r
e
ex
tr
ac
tio
n
p
r
o
ce
s
s
;
an
d
(
3
)
test
i
n
g
.
3
.
1
.
Da
t
a
co
llect
io
n
a
nd
t
e
m
pla
t
e
pro
du
ct
io
n
T
h
e
J
aw
i
ch
ar
ac
ter
s
w
er
e
co
llected
ac
co
r
d
in
g
to
th
e
f
o
llo
w
i
n
g
s
tep
s
:
(
1
)
w
r
ite
t
h
e
c
h
ar
ac
t
er
s
o
n
t
h
e
p
ap
er
;
(
2
)
s
ca
n
th
e
p
ap
er
an
d
s
av
e
t
h
e
ch
ar
ac
ter
i
m
a
g
e;
(
3
)
cr
o
p
t
h
e
ch
ar
ac
ter
p
ar
t
in
th
e
i
m
ag
e,
b
y
p
o
s
itio
n
i
n
g
t
h
e
cr
o
p
p
ed
p
ar
t
in
th
e
m
id
d
le
o
f
th
e
i
m
a
g
e;
an
d
(
4
)
n
o
r
m
alize
t
h
e
i
m
ag
es
i
n
to
1
6
0
×
1
4
0
p
ix
els.
T
h
e
J
a
w
i
c
h
ar
ac
ter
te
m
p
late
w
a
s
p
r
ep
ar
ed
in
t
w
o
f
o
r
m
s
:
b
in
ar
y
a
n
d
th
in
.
E
x
a
m
p
les
o
f
b
i
n
ar
y
te
m
p
late
J
a
w
i
c
h
ar
ac
ter
s
ar
e
s
h
o
w
n
i
n
in
Fig
u
r
e
1
,
w
h
ile
e
x
a
m
p
le
s
o
f
t
h
i
n
te
m
p
late
J
a
w
i
c
h
ar
ac
ter
s
ar
e
s
h
o
w
n
in
Fig
u
r
e
2
.
Fig
u
r
e
1
.
E
x
a
m
p
les o
f
b
in
ar
y
t
e
m
p
late
J
a
w
i c
h
ar
ac
ter
s
Fig
u
r
e
2
.
E
x
a
m
p
les o
f
th
i
n
te
m
p
late
J
a
w
i c
h
ar
ac
ter
s
3
.
2
.
F
ea
t
ure
ex
t
ra
ct
io
n pro
ce
s
s
I
n
th
e
f
ea
t
u
r
e
ex
tr
ac
tio
n
p
r
o
ce
s
s
,
(
1
)
w
as
u
s
ed
to
ca
lc
u
l
ate
all
m
o
m
en
t
v
al
u
es
o
f
th
e
ch
ar
ac
ter
i
m
a
g
es.
T
h
e
m
o
m
e
n
t
v
al
u
e
s
w
er
e
s
to
r
ed
an
d
u
s
ed
as
f
ea
t
u
r
e
in
r
ec
o
g
n
itio
n
p
r
o
ce
s
s
.
B
as
ed
o
n
th
ese
v
al
u
es,
th
e
T
R
alg
o
r
ith
m
r
ec
o
g
n
izes
th
e
ch
ar
ac
ter
s
.
T
h
e
m
o
m
en
t
v
alu
e
s
o
f
t
h
e
ch
ar
ac
ter
w
er
e
class
i
f
ied
u
s
in
g
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
.
3
.
3
.
T
esting
T
h
e
T
R
alg
o
r
ith
m
w
as
te
s
te
d
o
n
s
ets
o
f
r
o
tated
an
d
s
ca
led
ch
ar
ac
ter
s
,
to
f
in
d
o
u
t
wh
eth
er
t
h
e
alg
o
r
ith
m
ca
n
r
ec
o
g
n
ize
t
h
e
r
o
tated
an
d
s
ca
led
J
a
w
i
c
h
ar
ac
ter
s
.
T
h
e
ch
ar
ac
ter
s
w
er
e
r
o
tat
ed
9
0
o
an
d
1
8
0
o
to
th
e
r
ig
h
t
a
n
d
s
ca
led
b
y
s
ca
le
f
ac
to
r
s
o
f
2
an
d
0
.
5
.
F
o
u
r
teen
ch
ar
ac
ter
s
w
er
e
ch
o
s
e
n
,
an
d
ea
ch
ch
ar
ac
ter
w
a
s
ch
o
s
en
to
r
ep
r
esen
t
ea
ch
s
h
a
p
e
g
r
o
u
p
o
f
th
e
J
a
w
i
alp
h
ab
et.
W
e
u
s
ed
s
ev
e
n
c
h
ar
ac
ter
s
f
r
o
m
th
e
b
i
n
ar
y
te
m
p
late:
is
o
lated
j
i
m
,
en
d
-
f
o
r
m
h
a,
m
id
d
le
-
f
o
r
m
y
a,
i
n
itial
g
a,
in
itia
l
n
g
,
is
o
lated
p
a,
an
d
en
d
-
f
o
r
m
p
a,
an
d
s
ev
e
n
ch
ar
ac
ter
s
f
r
o
m
t
h
i
n
tem
p
late,
co
n
s
i
s
tin
g
o
f
is
o
lated
alif
,
is
o
lated
k
af
,
m
id
d
le
-
f
o
r
m
h
aa
,
m
id
d
le
-
f
o
r
m
y
a,
en
d
-
f
o
r
m
r
a,
m
id
d
le
-
f
o
r
m
h
a,
an
d
en
d
-
f
o
r
m
v
a.
Fu
r
t
h
er
m
o
r
e,
T
h
e
SVM
an
d
E
u
clid
ea
n
d
is
ta
n
ce
(
E
D)
as
class
i
f
ier
w
er
e
ap
p
lied
to
c
o
m
p
ar
e
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
.
W
e
ap
p
lied
th
e
SVM
cla
s
s
i
f
ier
a
n
d
E
D
to
t
h
e
m
o
m
e
n
t
v
al
u
es
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
Mo
men
t in
va
r
ia
n
t
-
b
a
s
ed
fe
a
tu
r
es fo
r
J
a
w
i c
h
a
r
a
cter reco
g
n
itio
n
(
F
itr
i A
r
n
ia
)
1715
ca
lcu
lated
b
y
(
1
)
,
an
d
th
en
co
m
p
ar
ed
th
e
r
es
u
lt
s
w
it
h
th
at
o
f
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
.
I
n
d
ata
tr
ain
in
g
f
o
r
SVM
an
d
E
D,
w
e
u
s
ed
f
ea
t
u
r
es
o
f
1
2
5
ch
ar
ac
ter
s
f
o
r
ea
c
h
b
in
ar
y
an
d
th
i
n
te
m
p
late.
I
n
test
i
n
g
s
t
ag
es,
w
e
u
s
ed
b
o
th
r
o
tated
an
d
s
ca
led
ch
ar
ac
ter
s
.
I
n
SVM
tr
ain
i
n
g
,
w
e
u
s
ed
o
n
e
ag
ain
s
t
all
s
tr
ateg
y
f
o
r
class
i
f
y
in
g
ea
ch
cla
s
s
o
f
ch
ar
ac
ter
.
Af
ter
tr
ain
i
n
g
,
w
e
test
ed
th
e
class
if
ied
clas
s
u
s
in
g
th
e
te
s
ti
n
g
c
h
ar
ac
ter
s
.
T
h
e
E
D
v
alu
e
w
a
s
co
m
p
u
ted
b
et
w
ee
n
t
h
e
f
ea
t
u
r
es,
i.e
.
,
t
h
e
m
o
m
e
n
t
v
al
u
e
s
o
f
b
i
n
ar
y
o
r
th
i
n
te
m
p
late
o
f
t
r
ain
in
g
an
d
te
s
ti
n
g
ch
ar
ac
ter
.
T
h
e
E
D
v
alu
e
w
as s
o
r
ted
,
an
d
th
e
s
m
alle
s
t E
D
v
al
u
e
in
d
icate
d
t
h
e
r
ec
o
g
n
ized
ch
ar
ac
ter
.
4.
RE
SU
L
T
S AN
D
AN
AL
Y
SI
S
T
h
e
T
R
alg
o
r
ith
m
clas
s
i
f
ied
t
h
e
J
a
w
i
c
h
ar
ac
ter
s
i
n
to
s
e
v
e
n
g
r
o
u
p
s
.
T
h
e
f
ir
s
t
g
r
o
u
p
co
n
s
is
ted
o
f
o
n
e,
i.e
.
,
th
e
f
ir
s
t
m
o
m
en
t
v
al
u
e.
I
n
th
is
ca
s
e,
t
h
e
ch
ar
ac
ter
is
r
ep
r
esen
ted
b
y
o
n
e,
i.e
.
,
th
e
f
i
r
s
t
m
o
m
en
t
v
al
u
es.
T
h
e
s
ec
o
n
d
g
r
o
u
p
co
n
s
is
ted
o
f
t
w
o
,
i.e
.
,
th
e
f
ir
s
t
a
n
d
t
h
e
s
e
co
n
d
m
o
m
en
t
v
alu
e
s
.
I
n
t
h
is
c
ase,
t
h
e
c
h
ar
ac
ter
is
r
ep
r
esen
ted
b
y
t
w
o
,
i.e
.
,
th
e
f
i
r
s
t
an
d
th
e
s
ec
o
n
d
m
o
m
e
n
ts
.
Si
m
i
lar
r
u
les
ar
e
ap
p
lied
to
th
e
r
em
ai
n
i
n
g
g
r
o
u
p
s
.
W
ith
in
t
h
e
s
a
m
e
g
r
o
u
p
,
ea
ch
c
h
ar
ac
ter
is
d
if
f
er
en
t
iated
b
y
a
u
n
iq
u
e
v
al
u
e
o
f
i
ts
h
ig
h
es
t
m
o
m
en
t.
4
.
1
.
Rec
o
g
nitio
n r
esu
lt
s
o
f
bin
a
ry
t
e
m
pla
t
e
cha
ra
ct
er
s
Fig
u
r
e
3
a
s
h
o
w
s
t
h
e
p
er
ce
n
tag
e
r
ec
o
g
n
itio
n
r
ate
i
n
ea
c
h
“
m
o
m
e
n
t
g
r
o
u
p
”
o
f
b
in
ar
y
te
m
p
late
ch
ar
ac
ter
s
.
B
ased
o
n
t
h
e
r
es
u
l
ts
,
t
h
er
e
w
as
o
n
e
c
h
ar
ac
ter
t
h
at
ca
n
b
e
r
ep
r
esen
ted
b
y
t
w
o
m
o
m
e
n
t,
i.e
.
,
0
.
8
%.
Fiv
e
c
h
ar
ac
ter
s
w
er
e
r
ep
r
esen
ted
b
y
t
h
r
ee
m
o
m
en
ts
,
o
r
4
%.
T
h
e
ch
ar
ac
ter
s
th
at
ca
n
b
e
r
ec
o
g
n
ized
b
y
f
o
u
r
m
o
m
e
n
t
s
w
er
e
15
ch
ar
ac
ter
s
o
r
1
2
%.
T
h
er
e
w
er
e
2
1
ch
ar
ac
ter
s
r
ec
o
g
n
izab
le
b
y
f
iv
e
m
o
m
e
n
t
s
,
o
r
1
6
.
8
%.
Fu
r
t
h
er
m
o
r
e,
r
ec
o
g
n
izab
le
c
h
ar
ac
ter
s
b
y
s
i
x
an
d
s
e
v
en
m
o
m
e
n
ts
w
er
e
3
2
ch
ar
ac
ter
s
(
2
5
.
6
%)
an
d
3
9
ch
ar
ac
ter
s
(
3
1
.
2
%),
r
esp
ec
tiv
el
y
.
Fro
m
a
to
tal
o
f
1
2
5
ch
ar
ac
ter
s
,
th
er
e
w
er
e
t
w
el
v
e
c
h
ar
ac
ter
s
th
at
ca
n
n
o
t
b
e
r
ec
o
g
n
ized
as
u
n
iq
u
e
ch
ar
ac
te
r
s
.
T
h
e
p
er
ce
n
tag
e
o
f
u
n
r
ec
o
g
n
izab
le
ch
ar
ac
ter
s
u
s
in
g
H
u
'
s
m
o
m
e
n
t
i
n
v
ar
ian
t
f
ea
t
u
r
es is
9
.
6
%
.
Fig
u
r
e
3b
s
h
o
w
s
t
h
e
r
ec
o
g
n
it
i
o
n
r
ate
o
f
th
i
n
te
m
p
la
te
ch
ar
a
cter
s
.
T
h
er
e
w
er
e
t
w
o
ch
ar
ac
t
er
s
th
at
ca
n
b
e
r
ep
r
esen
ted
b
y
t
w
o
m
o
m
en
ts
o
r
1
.
6
%.
T
h
e
p
er
ce
n
tag
e
o
f
r
ec
o
g
n
izab
le
ch
ar
ac
ter
s
b
ased
o
n
th
r
ee
m
o
m
en
t
s
w
a
s
4
.
8
%
o
r
6
ch
ar
ac
ter
s
f
r
o
m
t
h
e
to
tal
o
f
1
2
5
ch
ar
ac
ter
s
.
Up
to
1
4
ch
ar
ac
ter
s
o
r
1
1
.
2
%
an
d
3
0
ch
ar
ac
ter
s
o
r
24%
ca
n
b
e
r
ec
o
g
n
ized
b
y
f
o
u
r
an
d
f
i
v
e
m
o
m
en
ts
,
r
esp
ec
ti
v
el
y
.
F
u
r
th
er
m
o
r
e,
th
e
r
ec
o
g
n
ized
ch
ar
ac
ter
w
it
h
s
ix
m
o
m
e
n
ts
w
as
2
7
c
h
ar
ac
t
er
s
o
r
2
1
.
6
%,
as
th
e
r
ec
o
g
n
i
t
io
n
r
ate
w
i
th
s
e
v
en
m
o
m
e
n
ts
w
a
s
3
0
c
h
ar
ac
ter
s
(
2
4
%).
Six
teen
c
h
ar
ac
ter
s
w
er
e
n
o
t
r
ec
o
g
n
izab
le
as
u
n
iq
u
e
c
h
ar
ac
ter
s
,
w
h
ic
h
is
1
2
.
8
%
o
f
th
e
to
tal
o
f
1
2
5
ch
ar
ac
ter
s
.
(
a)
(
b
)
Fig
u
r
e
3
.
P
er
ce
n
tag
e
o
f
r
ec
o
g
n
i
tio
n
i
n
ea
c
h
“m
o
m
en
t
g
r
o
u
p
”
o
f
:
(
a)
b
in
ar
y
te
m
p
late
ch
ar
ac
ter
s
,
(
b
)
th
in
te
m
p
late
c
h
ar
ac
ter
s
4
.
2
.
T
esting
w
it
h
ro
t
a
t
ed
a
nd
s
ca
led c
ha
ra
ct
er
s
T
ab
les
3
an
d
T
a
b
le
4
s
h
o
w
t
h
e
tes
t
r
es
u
lts
w
it
h
b
in
ar
y
a
n
d
th
i
n
ch
ar
ac
ter
s
,
in
cl
u
d
i
n
g
r
o
tated
an
d
s
ca
led
ch
ar
ac
ter
s
.
Se
v
e
n
c
h
ar
ac
ter
s
w
er
e
ch
o
s
e
n
as
th
e
te
s
tin
g
c
h
ar
ac
ter
,
an
d
s
h
o
w
n
i
n
‘
in
p
u
t
c
h
ar
ac
ter
’
co
lu
m
n
.
E
ac
h
c
h
ar
ac
ter
w
as
r
o
tated
b
y
9
0
o
an
d
1
8
0
o
an
d
w
as
s
ca
led
b
y
t
w
o
s
ca
li
n
g
f
ac
to
r
:
0
.
5
an
d
2
.
T
h
e
u
n
r
ec
o
g
n
ized
ch
ar
ac
ter
s
o
f
b
in
ar
y
te
m
p
late
as
s
h
o
w
n
i
n
T
ab
le
3
a
r
e
J
im
in
itial
-
f
o
r
m
(
r
o
t
ated
1
8
0
o
)
,
Ha
en
d
-
f
o
r
m
(
r
o
tated
1
8
0
o
)
,
P
en
d
-
f
o
r
m
(
r
o
tated
9
0
o
)
,
G
in
itial
-
f
o
r
m
(
0
.
5
f
ac
to
r
s
ca
led
)
,
an
d
Ng
i
n
it
ial
-
f
o
r
m
(
0
.
5
f
ac
to
r
s
ca
led
)
.
T
ab
le
3
s
h
o
w
s
th
at
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
ca
n
r
ec
o
g
n
ize
ele
v
en
r
o
tated
an
d
t
w
el
v
e
s
ca
led
ch
ar
ac
ter
s
f
r
o
m
to
tal
2
8
ch
ar
ac
ter
s
.
As s
h
o
w
n
in
T
ab
le
3
,
ch
ar
ac
ter
J
im
in
in
itial
f
o
r
m
h
a
s
d
if
f
er
en
t
m
o
m
e
n
t
v
al
u
e
f
o
r
ea
ch
s
ca
led
f
ac
to
r
;
th
e
ch
ar
ac
ter
ـج
w
it
h
s
ca
led
f
ac
to
r
o
f
0
.
5
h
as
th
e
m
o
m
e
n
t
s
1
,
4
,
1
2
,
1
0
,
2
1
,
.
.
.
,
w
h
ile
t
h
e
ch
ar
ac
ter
ـج
w
it
h
s
ca
led
f
ac
to
r
o
f
2
h
a
s
t
h
e
m
o
m
e
n
ts
1
,
4
,
1
2
,
1
0
,
2
2
,
.
.
.
.
H
o
w
e
v
er
,
b
ec
au
s
e
t
h
e
m
o
m
en
t
v
alu
e
o
f
ea
c
h
s
ca
led
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
3
,
J
u
n
e
2
0
1
9
:
1
7
1
1
-
1
7
1
9
1716
ch
ar
ac
ter
m
atc
h
es
t
h
e
s
av
e
d
f
ea
tu
r
e,
th
at
w
a
s
th
e
f
ir
s
t
f
o
u
r
m
o
m
e
n
ts
,
1
,
4
,
1
2
,
1
0
,
th
e
ch
ar
ac
ter
w
a
s
r
ec
o
g
n
ized
.
I
n
ad
d
itio
n
,
T
ab
le
4
s
h
o
w
s
t
h
at
Kaf
i
s
o
lated
-
f
o
r
m
,
Haa
m
id
d
le
-
f
o
r
m
,
R
e
n
d
-
f
o
r
m
,
Ya
m
i
d
d
le
-
f
o
r
m
,
Ha
en
d
-
f
o
r
m
,
V
e
n
d
-
f
o
r
m
,
a
n
d
alif
is
o
lated
-
f
o
r
m
ar
e
th
e
u
n
r
ec
o
g
n
ized
ch
ar
ac
ter
s
o
f
t
h
i
n
te
m
p
la
te.
T
ab
le
4
p
r
esen
ts
n
in
e
r
o
tated
an
d
ei
g
h
t
s
ca
led
ch
ar
ac
ter
s
f
o
r
m
2
8
ch
ar
ac
ter
s
s
u
cc
ee
d
to
b
e
r
ec
o
g
n
i
ze
d
.
T
h
ese
r
esu
lt
s
s
u
g
g
e
s
t
th
at
th
e
T
R
alg
o
r
i
th
m
i
s
r
o
b
u
s
t
to
r
o
tatio
n
an
d
s
ca
li
n
g
o
p
er
atio
n
s
,
a
s
w
il
l
b
e
d
is
c
u
s
s
ed
in
th
e
f
o
llo
w
i
n
g
.
T
ab
le
3
.
T
R
A
lg
o
r
it
h
m
T
esti
n
g
o
n
B
in
ar
y
C
h
ar
ac
ter
s
T
ab
le
4
.
T
R
A
lg
o
r
it
h
m
T
esti
n
g
o
n
T
h
in
C
h
ar
ac
ter
s
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
Mo
men
t in
va
r
ia
n
t
-
b
a
s
ed
fe
a
tu
r
es fo
r
J
a
w
i c
h
a
r
a
cter reco
g
n
itio
n
(
F
itr
i A
r
n
ia
)
1717
An
o
th
er
ex
a
m
p
le
i
n
T
ab
le
4
,
ch
ar
ac
ter
ـي
ـ
(
Ya
m
id
d
le
-
f
o
r
m
)
w
a
s
r
o
tated
b
y
9
0
◦
,
an
d
h
as
th
e
m
o
m
e
n
t
v
alu
e
1
,
5
,
2
1
,
2
0
,
2
3
,
4
2
,
an
d
ch
ar
ac
ter
ـي
ـ
w
a
s
r
o
tated
b
y
1
8
0
◦
h
as
th
e
m
o
m
en
t
v
al
u
e
1
,
5
,
2
1
,
2
0
,
2
3
,
4
1
,
th
u
s
b
o
th
ch
ar
ac
ter
s
ar
e
r
ec
o
g
n
ize
d
as
ـ
ي
ـ
,
b
ec
au
s
e
th
e
f
ea
t
u
r
e
ar
e
1
,
5
,
2
1
,
2
0
.
T
h
e
d
if
f
er
en
t
m
o
m
e
n
t
v
alu
e
at
t
h
e
f
i
f
th
to
s
e
v
en
th
m
o
m
e
n
t
is
i
g
n
o
r
ed
,
b
ec
au
s
e
it
w
a
s
n
o
t
s
av
ed
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
r
esu
lted
in
b
etter
r
ec
o
g
n
itio
n
r
ate
o
n
b
in
ar
y
te
m
p
late
t
h
an
t
h
i
n
te
m
p
la
te.
Mo
s
t
ch
ar
ac
ter
s
i
n
b
in
ar
y
te
m
p
la
t
e
b
ec
o
m
e
u
n
iq
u
e
at
th
e
s
i
x
th
an
d
s
e
v
e
n
t
h
m
o
m
e
n
t
s
w
h
ile
m
o
s
t
c
h
ar
ac
ter
s
i
n
th
i
n
te
m
p
late
b
ec
o
m
e
u
n
iq
u
e
i
n
th
e
f
i
f
t
h
,
t
h
e
s
i
x
th
,
an
d
th
e
s
ev
e
n
t
h
m
o
m
e
n
ts
.
T
ab
le
3
an
d
4
d
em
o
n
s
tr
ates
t
h
at
th
e
T
R
alg
o
r
ith
m
ca
n
r
ec
o
g
n
ize
th
e
r
o
tated
an
d
s
ca
led
ch
ar
ac
ter
s
,
alth
o
u
g
h
t
h
ese
ch
ar
ac
ter
s
h
a
v
e
t
h
e
d
if
f
er
en
ce
m
o
m
en
t
v
al
u
es
at
t
h
e
la
test
m
o
m
e
n
ts
.
T
h
e
f
ea
t
u
r
es c
o
n
tai
n
n
o
t a
ll
m
o
m
e
n
t v
al
u
e
co
m
b
i
n
atio
n
s
,
b
u
t o
n
l
y
th
e
u
n
iq
u
e
o
n
e
s
.
4
.
3
.
Co
m
pa
riso
n w
it
h o
t
her
cla
s
s
if
ica
t
io
n
m
et
ho
ds
I
n
t
h
is
s
u
b
s
ec
tio
n
,
w
e
p
r
ese
n
t
co
m
p
ar
is
o
n
r
es
u
lt
o
f
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
w
ith
o
th
er
b
en
ch
m
ar
k
i
n
g
class
i
f
icatio
n
m
et
h
o
d
s
.
So
m
e
o
f
t
h
e
b
en
c
h
m
ar
k
in
g
p
atter
n
c
lass
i
f
ier
s
ar
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
an
d
E
u
clid
ea
n
D
is
ta
n
ce
(
E
D)
.
T
ab
le
5
s
h
o
w
s
th
e
r
es
u
lt
s
o
f
r
ec
o
g
n
i
tio
n
u
s
i
n
g
SV
M
clas
s
if
ier
an
d
E
D.
T
h
e
SVM
class
i
f
ier
u
s
ed
1
2
5
ch
ar
ac
ter
s
in
d
at
a
tr
ain
i
n
g
p
h
ase.
B
y
u
s
i
n
g
b
i
n
ar
y
te
m
p
late,
SVM
f
ail
ed
to
r
ec
o
g
n
ize
t
h
e
J
a
w
i
c
h
ar
ac
ter
.
No
n
e
o
f
te
s
ti
n
g
c
h
ar
ac
ter
w
as
r
ec
o
g
n
ized
,
t
h
u
s
,
t
h
e
r
ec
o
g
n
itio
n
ac
c
u
r
ac
y
o
f
b
i
n
ar
y
te
m
p
late
w
a
s
0
%.
B
y
u
s
i
n
g
t
h
in
te
m
p
late,
SVM
r
ec
o
g
n
ized
t
h
r
ee
ch
ar
ac
ter
s
,
t
h
u
s
,
th
e
r
ec
o
g
n
itio
n
ac
c
u
r
ac
y
w
a
s
1
0
.
7
1
%.
T
h
e
SVM
h
ad
b
etter
p
er
f
o
r
m
an
ce
o
n
t
h
i
n
ch
ar
ac
ter
s
th
an
b
i
n
ar
y
ch
ar
ac
ter
s
.
On
th
e
o
th
er
h
a
n
d
,
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
h
ad
b
etter
r
ec
o
g
n
itio
n
r
ate
o
n
b
in
ar
y
ch
ar
ac
ter
s
th
an
t
h
i
n
ch
ar
ac
ter
s
.
E
D
ac
h
ie
v
ed
b
etter
r
es
u
lt
t
h
a
n
S
VM
cla
s
s
i
f
ier
b
u
t
le
s
s
th
a
n
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
.
B
y
u
s
in
g
b
i
n
ar
y
te
m
p
late,
E
D
r
ec
o
g
n
ized
ele
v
en
ch
ar
ac
ter
s
,
th
u
s
,
th
e
r
ec
o
g
n
itio
n
ac
cu
r
ac
y
w
as
3
9
.
2
9
%.
On
t
h
i
n
te
m
p
late,
E
D
h
ad
less
ac
c
u
r
ate
th
a
n
b
i
n
ar
y
te
m
p
late
b
y
r
ec
o
g
n
izi
n
g
4
ch
ar
ac
ter
s
o
r
1
4
.
2
9
%.
T
r
a
in
i
n
g
p
h
as
e
is
n
o
t
r
eq
u
ir
ed
b
y
t
h
e
p
r
o
p
o
s
ed
an
d
E
D
m
et
h
o
d
,
in
s
tead
b
o
th
m
eth
o
d
s
d
ir
ec
tly
r
ec
o
g
n
ize
t
h
e
ch
ar
ac
ter
s
b
y
m
atc
h
i
n
g
th
e
f
ea
tu
r
es,
i.e
.
,
th
e
m
o
m
e
n
t
v
alu
e
s
o
f
ea
c
h
ch
ar
ac
ter
.
T
ab
le
5
.
C
o
m
p
ar
is
o
n
o
f
P
r
o
p
o
s
ed
Me
th
o
d
w
i
th
S
VM
an
d
E
D
C
l
a
ssi
f
i
e
r
T
r
a
i
n
i
n
g
c
h
a
r
a
c
t
e
r
T
e
st
i
n
g
c
h
a
r
a
c
t
e
r
R
e
c
o
g
n
i
z
e
d
A
c
c
u
r
a
c
y
B
i
n
a
r
y
t
e
mp
l
a
t
e
S
V
M
1
2
5
28
0
0%
ED
-
28
11
3
9
.
2
9
%
P
r
o
p
o
se
d
M
e
t
h
o
d
-
28
23
8
2
.
1
4
%
T
h
i
n
t
e
mp
l
a
t
e
S
V
M
1
2
5
28
3
1
0
.
7
1
%
ED
-
28
4
1
4
.
2
9
%
P
r
o
p
o
se
d
M
e
t
h
o
d
-
28
17
6
0
.
7
1
%
SVM
clas
s
i
f
ier
h
a
s
lo
w
er
r
ec
o
g
n
i
tio
n
ac
c
u
r
ac
y
.
C
o
m
p
ar
i
n
g
t
o
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
an
d
E
D
th
at
d
id
n
o
t
r
eq
u
ir
e
m
o
r
e
d
ataset
to
p
er
f
o
r
m
tr
ain
in
g
p
r
o
ce
s
s
,
SVM
f
ailed
in
r
ec
o
g
n
iz
in
g
J
a
w
i
ch
ar
ac
ter
w
it
h
o
n
l
y
o
n
e
d
ataset.
T
h
is
lo
w
er
p
er
f
o
r
m
a
n
ce
w
a
s
ca
u
s
ed
b
y
t
h
e
l
ac
k
o
f
d
ata
tr
ai
n
i
n
g
f
o
r
SVM
cla
s
s
i
f
ier
.
I
n
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
,
th
e
u
s
ed
o
f
l
ess
tr
ain
i
n
g
d
ata
w
o
u
ld
d
ec
r
ea
s
e
th
e
m
eth
o
d
’
s
p
er
f
o
r
m
a
n
ce
.
On
th
e
o
t
h
er
h
a
n
d
,
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
w
as
n
o
t
af
f
ec
ted
b
y
t
h
e
n
u
m
b
er
o
f
tr
ai
n
in
g
d
ata
b
ec
au
s
e
i
t
d
id
n
o
t
n
ee
d
tr
ai
n
in
g
s
ta
g
es,
an
d
u
s
e
t
h
e
f
ea
tu
r
es d
ir
ec
tl
y
i
n
r
ec
o
g
n
itio
n
p
r
o
ce
s
s
.
5.
CO
NCLU
SI
O
N
I
n
t
h
is
ar
ticle,
w
e
h
a
v
e
p
r
es
en
ted
a
n
o
v
el
f
ea
tu
r
e
e
x
tr
ac
t
io
n
an
d
r
ec
o
g
n
i
tio
n
tech
n
iq
u
e
f
o
r
J
a
w
i
ch
ar
ac
ter
s
b
ased
o
n
m
o
m
en
t
i
n
v
ar
ia
n
t
f
ea
t
u
r
es,
w
h
ic
h
i
s
ca
l
led
th
e
T
r
ee
R
o
o
t
(
T
R
)
alg
o
r
it
h
m
.
T
h
is
al
g
o
r
ith
m
ass
i
g
n
ed
ea
c
h
c
h
ar
ac
ter
a
u
n
iq
u
e
co
m
b
i
n
atio
n
o
f
H
u
’
s
m
o
m
en
t
v
al
u
es.
T
h
e
T
R
al
g
o
r
ith
m
w
a
s
ap
p
lied
to
t
wo
ch
ar
ac
ter
te
m
p
late
s
,
b
in
ar
y
a
n
d
th
i
n
,
ea
ch
co
n
s
i
s
ti
n
g
o
f
1
2
5
J
a
w
i
c
h
ar
ac
ter
s
.
I
n
t
h
e
b
in
ar
y
te
m
p
late,
t
h
e
T
R
ca
n
r
ec
o
g
n
iz
e
9
0
.
4
%
o
f
t
h
e
c
h
ar
ac
ter
s
,
w
h
ile
i
n
t
h
e
t
h
in
te
m
p
late
it
ca
n
r
ec
o
g
n
ize
8
7
.
2
%
o
f
t
h
e
c
h
ar
ac
ter
s
.
T
h
e
r
esu
lts
s
h
o
w
ed
th
at
t
h
e
T
R
alg
o
r
ith
m
s
u
cc
es
s
f
u
ll
y
r
ec
o
g
n
ized
r
o
tated
an
d
s
ca
led
ch
ar
ac
ter
s
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
ac
h
iev
ed
h
i
g
h
er
r
ec
o
g
n
itio
n
r
ate,
co
m
p
ar
ed
w
ith
S
VM
class
if
ier
an
d
E
D
u
s
i
n
g
t
h
e
s
a
m
e
test
i
n
g
c
h
ar
ac
ter
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
is
r
esear
ch
is
f
u
n
d
ed
b
y
S
y
i
ah
Ku
ala
Un
i
v
er
s
it
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
3
,
J
u
n
e
2
0
1
9
:
1
7
1
1
-
1
7
1
9
1718
RE
F
E
R
E
NC
E
S
[1
]
M
o
a
in
A
J.
“
L
a
n
g
u
a
g
e
d
e
si
g
n
;
Histo
ry
o
f
Ja
v
a
n
e
se
c
h
a
ra
c
ter
(
in
Ba
h
a
sa
)
,”
De
wa
n
Ba
h
a
s
a
d
a
n
Pu
st
a
k
a
;
1
9
9
6
.
[2
]
K.
S
a
d
d
a
m
i,
e
t
a
l
.
,
“
A
Da
tab
a
s
e
o
f
P
rin
ted
Ja
w
i
Ch
a
ra
c
t
e
r
I
m
a
g
e
,”
2
0
1
5
IEE
E
Co
n
fer
e
n
c
e
o
n
Ima
g
e
In
fo
rm
a
ti
o
n
Pro
c
e
ss
in
g
(
ICIIP
),
p
p
.
5
6
-
5
9
,
2
0
1
5
.
[3
]
M
.
F
.
Na
sru
d
in
,
e
t
a
l
.
,
“
Ha
n
d
w
rit
t
e
n
c
u
rsiv
e
Ja
w
i
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
:
A
su
rv
e
y
,”
IEE
E
Co
n
fer
e
n
c
e
o
n
Co
m
p
u
ter
Gr
a
p
h
ics
,
Im
a
g
i
n
g
a
n
d
V
isu
a
li
sa
t
io
n
(
CGIV)
,
p
p
.
2
4
7
-
2
5
6
,
2
0
0
8
.
[4
]
F
.
A
rn
ia
a
n
d
K.
M
u
n
a
d
i,
“
Bin
a
ri
z
a
ti
o
n
o
f
A
n
c
ien
t
Do
c
u
m
e
n
t
Im
a
g
e
s
b
a
se
d
o
n
M
u
lt
ip
e
a
k
Histo
g
ra
m
A
ss
u
m
p
ti
o
n
,
”
T
e
lec
o
mm
u
n
ica
ti
o
n
Co
mp
u
ti
n
g
E
lec
tro
n
ics
a
n
d
C
o
n
tr
o
l
(
T
EL
KOM
NIKA
)
.
v
o
l.
1
5
(
3
),
p
p
.
1
3
1
7
-
1
3
2
7
,
2
0
1
7
.
[5
]
T
.
N.
Ya
n
g
a
n
d
S
.
D.
W
a
n
g
.
“
A
ro
tatio
n
in
v
a
rian
t
p
rin
te
d
Ch
in
e
se
c
h
a
ra
c
ter
r
e
c
o
g
n
it
io
n
sy
ste
m
,”
Pa
tt
e
rn
re
c
o
g
n
it
i
o
n
letter
s
,
v
o
l.
2
2
(
2
),
p
p
.
8
5
-
9
5
,
2
0
0
1
.
[6
]
N.N.
Kh
a
rm
a
a
n
d
R.
K.
W
a
rd
,
“
A
n
o
v
e
l
in
v
a
rian
t
m
a
p
p
in
g
a
p
p
li
e
d
to
h
a
n
d
-
w
rit
ten
A
ra
b
ic
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
,”
Pa
tt
e
rn
Rec
o
g
n
it
io
n
,
v
o
l
.
3
4
(1
1
),
p
p
.
2
1
1
5
-
2
1
2
0
,
2
0
0
1
.
[7
]
A
.
H.
H
a
ss
in
,
“
P
rin
ted
A
ra
b
ic
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
u
si
n
g
HMM
,”
J
o
u
rn
a
l
o
f
Co
mp
u
ter
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
.
V
o
l
.
1
9
(
4
),
p
p
.
5
3
8
-
4
3
,
2
0
0
4
.
[8
]
R.
S
.
Ku
n
te
a
n
d
R.
D.
S
a
m
u
e
l,
“
A
si
m
p
le
a
n
d
e
ff
icie
n
t
o
p
ti
c
a
l
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
sy
ste
m
f
o
r
b
a
sic
s
y
m
b
o
ls
in
p
rin
te
d
Ka
n
n
a
d
a
tex
t,
”
S
a
d
h
a
n
a
,
v
o
l.
3
2
(5
),
2
0
0
7
.
[9
]
Iq
b
a
l
A
,
A
BM
M
,
T
a
h
sin
A
,
S
a
tt
a
r
M
A
,
Isla
m
M
M
,
M
u
ra
se
K.
“
A
No
v
e
l
A
lg
o
rit
h
m
f
o
r
T
ra
n
sla
ti
o
n
,
R
o
tatio
n
a
n
d
S
c
a
le In
v
a
rian
t
Ch
a
ra
c
ter Rec
o
g
n
it
io
n
,”
Ja
p
a
n
S
o
c
iety
f
o
r
F
u
z
z
y
Th
e
o
ry
a
n
d
In
telli
g
e
n
t
I
n
f
o
rm
a
ti
c
s.
In
S
CIS
&
IS
I
S
S
CIS
&
IS
IS
2
0
0
8
(p
p
.
1
3
6
7
-
1
3
7
2
),
2
0
0
8
.
[1
0
]
F
.
A
rn
ia,
e
t
a
l
.
,
“
I
m
p
ro
v
e
m
e
n
t
o
f
b
in
a
riza
ti
o
n
p
e
rf
o
rm
a
n
c
e
b
y
a
p
p
ly
in
g
d
c
t
a
s
p
re
-
p
ro
c
e
ss
in
g
p
ro
c
e
d
u
r
e
”
6
th
In
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m
Co
mm
u
n
ic
a
ti
o
n
s,
Co
n
tro
l
a
n
d
S
i
g
n
a
l
Pro
c
e
ss
in
g
(
IS
CCS
P)
,
2
0
1
4
o
n
2
0
1
4
M
a
y
2
1
(p
p
.
1
2
8
-
1
3
2
),
2
0
1
4
.
[1
1
]
S
.
M
u
c
h
a
ll
il
,
e
t
a
l
.
,
“
P
e
rf
o
rm
a
n
c
e
Co
m
p
a
riso
n
Of
D
e
n
o
isin
g
M
e
th
o
d
s
F
o
r
Histo
rica
l
Do
c
u
m
e
n
ts
,”
J
u
rn
a
l
T
e
k
n
o
lo
g
i
,
v
o
l.
7
7
,
n
o
.
2
2
,
p
p
.
1
4
3
-
1
3
7
,
2
0
1
5
.
[1
2
]
F
a
rd
ian
,
e
t
a
l.
,
“
I
d
e
n
ti
f
ica
ti
o
n
o
f
M
o
st
S
u
it
a
b
le
Bi
n
a
risa
ti
o
n
M
e
th
o
d
F
o
r
A
c
e
h
n
e
se
A
n
c
ie
n
t
M
a
n
u
sc
rip
ts
Re
sto
ra
ti
o
n
S
o
f
tw
a
re
Us
e
r
G
u
id
e
,”
J
u
rn
a
l
T
e
k
n
o
l
o
g
i
,
v
o
l.
7
7
,
n
o
.
2
2
,
p
p
.
9
5
-
1
0
2
,
2
0
1
5
.
[1
3
]
M
.
F
.
Na
sru
d
in
,
e
t
a
l
.
,
“
In
v
a
rian
t
f
e
a
tu
re
s
f
ro
m
th
e
trac
e
tran
sf
o
r
m
f
o
r
ja
w
i
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
,
”
In
ter
n
a
ti
o
n
a
l
W
o
rk
-
Co
n
fer
e
n
c
e
o
n
Arti
fi
c
ia
l
Ne
u
ra
l
Ne
tw
o
rk
s
,
p
p
.
2
5
6
-
2
6
3
,
2
0
0
9
.
[1
4
]
N.S
.
Ra
n
i
,
e
t
a
l.
,
“
A
Zo
n
e
Ba
se
d
A
p
p
ro
a
c
h
f
o
r
Clas
sif
ica
ti
o
n
a
n
d
Re
c
o
g
n
it
io
n
o
f
T
e
lu
g
u
Ha
n
d
w
rit
t
e
n
Ch
a
ra
c
ters
,”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
E
lec
trica
l
a
n
d
C
o
mp
u
ter
En
g
in
e
e
rin
g
,
V
o
l
.
6
(
4
),
p
p
.
1
6
4
7
,
2
0
1
6
.
[1
5
]
M
.
Ke
f
,
e
t
a
l.
,
“
A
n
o
v
e
l
f
u
z
z
y
a
p
p
ro
a
c
h
f
o
r
h
a
n
d
w
rit
ten
A
ra
b
ic
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
,”
Pa
tt
e
r
n
An
a
lys
is
a
n
d
Ap
p
li
c
a
ti
o
n
s
,
v
o
l
.
1
9
(4
),
p
p
.
1
0
4
1
-
1
0
5
6
,
2
0
1
6
.
[1
6
]
M
.
F
.
Na
sru
d
in
a
n
d
M
.
P
e
tro
u
,
“
Off
li
n
e
h
a
n
d
w
rit
ten
Ja
w
i
re
c
o
g
n
it
io
n
u
si
n
g
th
e
trac
e
tran
sf
o
rm
,
”
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Pa
tt
e
rn
A
n
a
lys
is
a
n
d
I
n
telli
g
e
n
t
R
o
b
o
ti
c
s (
ICPA
IR)
,
p
p
.
8
7
-
9
1
,
2
0
1
1
.
[1
7
]
Z.
Ra
z
a
k
,
e
t
a
l.
,
“
O
ff
-
li
n
e
ja
w
i
h
a
n
d
w
rit
in
g
re
c
o
g
n
it
io
n
u
si
n
g
h
a
m
m
in
g
c
las
si
f
ica
ti
o
n
,
”
In
fo
rm
a
t
io
n
T
e
c
h
n
o
l
o
g
y
J
o
u
rn
a
l
,
v
o
l
.
8
,
p
p
.
9
7
1
-
8
1
,
2
0
0
9
.
[1
8
]
S
.
Ha
y
k
in
,
“
Ne
u
ra
l
n
e
tw
o
rk
s: a c
o
m
p
re
h
e
n
siv
e
f
o
u
n
d
a
ti
o
n
,”
2
nd
e
d
.
Ne
w
Je
r
se
y
:
Pre
n
ti
c
e
Ha
ll
PT
R
;
1
9
9
9
.
[1
9
]
S
.
Ku
ti
n
a
n
d
P
.
Niy
o
g
i,
“
T
h
e
in
tera
c
ti
o
n
o
f
sta
b
il
it
y
a
n
d
w
e
a
k
n
e
s
s
in
A
d
a
Bo
o
st
,
”
T
e
c
h
n
ica
l
Rep
o
r
t
,
T
R
-
2
0
0
1
–
3
0
,
2
0
0
1
.
[2
0
]
M
.
F
.
Na
sru
d
in
,
e
t
a
l.
,
“
Ob
jec
t
sig
n
a
tu
re
f
e
a
tu
re
s
se
lec
ti
o
n
fo
r
h
a
n
d
w
rit
ten
Ja
w
i
r
e
c
o
g
n
it
io
n
,
”
Distrib
u
ted
Co
mp
u
t
in
g
a
n
d
Arti
fi
c
i
a
l
I
n
telli
g
e
n
c
e
,
p
p
.
6
8
9
-
9
8
,
2
0
1
0
.
[2
1
]
A
.
H
e
r
y
a
n
to
,
e
t
a
l.
,
“
O
ff
li
n
e
Ja
w
i
h
a
n
d
w
rit
ten
re
c
o
g
n
ize
r
u
sin
g
h
y
b
rid
a
rti
f
icia
l
n
e
u
ra
l
n
e
t
w
o
rk
s
a
n
d
d
y
n
a
m
ic
p
ro
g
ra
m
m
in
g
,
”
In
ter
n
a
ti
o
n
a
l
S
y
mp
o
siu
m
in
I
n
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
v
o
l.
2
,
p
p
.
1
-
6
,
2
0
0
8
.
[2
2
]
R.
Re
d
ik
a
,
e
t
a
l.
,
“
Ha
n
d
w
rit
ten
Ja
w
i
w
o
rd
s
re
c
o
g
n
it
io
n
u
sin
g
h
id
d
e
n
M
a
rk
o
v
m
o
d
e
ls
,”
In
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m
in
In
fo
rm
a
t
io
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
2
,
p
p
.
1
-
5
,
2
0
0
8
.
[2
3
]
M
.
F
.
Na
sru
d
i
n
,
e
t
a
l.
,
Ja
w
i
c
h
a
ra
c
ter
re
c
o
g
n
it
io
n
u
sin
g
th
e
tra
c
e
tran
s
f
o
rm
,”
7
th
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
in
Co
mp
u
ter
Gr
a
p
h
ics
,
Ima
g
in
g
a
n
d
Vi
su
a
li
za
ti
o
n
(
CGIV)
,
p
p
.
1
5
1
-
1
5
6
,
2
0
1
0
.
[2
4
]
M
.
Ch
e
riet,
e
t
a
l.
,
“
Ch
a
ra
c
ter
re
c
o
g
n
it
io
n
sy
ste
m
s:
a
g
u
id
e
f
o
r
stu
d
e
n
ts
a
n
d
p
ra
c
ti
ti
o
n
e
rs
,”
Ne
w
Je
r
se
y
:
J
o
h
n
W
il
e
y
&
S
o
n
s
,
2
0
0
7
.
[2
5
]
V
.
V
a
p
n
ik
,
“
T
h
e
n
a
tu
re
of
sta
ti
stica
l
lea
rn
in
g
th
e
o
ry
,
”
2
nd
e
d
.
Ne
w
Yo
rk
:
S
p
ri
n
g
e
r sc
ien
c
e
&
b
u
sin
e
ss
me
d
ia
;
2
0
1
3
.
[2
6
]
M
.
G
h
ias
si,
e
t
a
l.
,
“
A
u
to
m
a
t
e
d
tex
t
c
las
si
f
ica
ti
o
n
u
sin
g
a
d
y
n
a
m
ic
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
m
o
d
e
l
,”
Exp
e
rt
S
y
ste
ms
wit
h
A
p
p
l
ica
ti
o
n
s
,
v
o
l.
3
9
(
1
2
),
p
p
.
1
0
9
6
7
-
1
0
9
7
6
,
2
0
1
2
.
[2
7
]
S
.
A
ro
ra
,
e
t
a
l.
,
“
P
e
rf
o
rm
a
n
c
e
c
o
m
p
a
riso
n
o
f
S
V
M
a
n
d
A
NN
f
o
r
h
a
n
d
w
rit
ten
d
e
v
n
a
g
a
ri
c
h
a
ra
c
te
r
re
c
o
g
n
it
io
n
,
”
a
rXiv p
re
p
ri
n
t
a
rX
iv
:1
0
0
6
.
5
9
0
2
,
2
0
1
0
.
[2
8
]
Z.
Ju
a
n
d
C.
S
u
,
“
On
e
-
p
a
ss
M
o
m
e
n
t
A
l
g
o
rit
h
m
f
o
r
G
ra
p
h
ica
l
P
rim
it
iv
e
s
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
i
n
e
e
rin
g
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
1
2
(5
),
p
p
.
3
8
1
8
-
3
8
2
4
,
2
0
1
4
.
[2
9
]
R.
M
u
ra
li
d
h
a
ra
n
a
n
d
C.
Ch
a
n
d
ra
se
k
a
r,
“
Ob
jec
t
re
c
o
g
n
it
io
n
u
s
in
g
S
V
M
-
KN
N
b
a
se
d
o
n
g
e
o
m
e
tri
c
m
o
m
e
n
t
in
v
a
rian
t
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
m
p
u
ter
T
re
n
d
s a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
(
1
),
p
p
.
2
1
5
-
2
2
0
,
2
0
1
1
.
[3
0
]
M
.
K.
Hu
,
“
V
is
u
a
l
p
a
tt
e
rn
re
c
o
g
n
it
io
n
b
y
m
o
m
e
n
t
in
v
a
rian
ts
,
”
IRE
tra
n
sa
c
ti
o
n
s
o
n
in
f
o
rm
a
ti
o
n
th
e
o
ry
,
v
o
l.
8
(2
)
,
p
p
.
1
7
9
-
8
7
,
1
9
6
2
.
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
Mo
men
t in
va
r
ia
n
t
-
b
a
s
ed
fe
a
tu
r
es fo
r
J
a
w
i c
h
a
r
a
cter reco
g
n
itio
n
(
F
itr
i A
r
n
ia
)
1719
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
Fi
tr
i
Ar
n
ia
re
c
e
i
v
e
d
B.
En
g
d
e
g
re
e
f
ro
m
Un
iv
e
rsitas
S
u
m
a
ter
a
Uta
ra
(USU),
M
e
d
a
n
in
1
9
9
7
.
S
h
e
f
in
ish
e
d
h
e
r
m
a
ste
r
a
n
d
d
o
c
to
ra
l
d
e
g
re
e
f
ro
m
Un
iv
e
rss
it
y
o
f
Ne
w
S
o
u
th
W
a
les
(UN
S
W
),
S
y
d
n
e
y
,
A
u
stra
li
a
a
n
d
T
o
k
y
o
M
e
tro
p
o
li
tan
U
n
iv
e
rsity
,
J
a
p
a
n
in
2
0
0
4
a
n
d
2
0
0
8
re
sp
e
c
ti
v
e
l
y
.
S
h
e
h
a
s
b
e
e
n
w
it
h
th
e
De
p
a
rtm
e
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
,
S
y
iah
Ku
a
la
Un
iv
e
rsit
y
sin
c
e
1
9
9
9
.
Dr.
A
rn
ia
w
a
s
a
v
isit
in
g
s
c
h
o
lar
in
T
o
k
y
o
M
e
tro
p
o
li
ta
n
Un
iv
e
rsity
(T
M
U),
T
o
k
y
o
,
Ja
p
a
n
in
2
0
1
3
a
n
d
S
u
ley
m
a
n
De
m
irel
Un
iv
e
rsi
ty
(S
DU
),
Isp
a
rta,
T
u
rk
e
y
in
2
0
1
7
.
S
h
e
is
a
m
e
m
b
e
r
o
f
IEE
E
a
n
d
A
P
S
I
P
A
.
He
r
re
se
a
rc
h
in
tere
sts
a
re
sig
n
a
l,
i
m
a
g
e
a
n
d
m
u
lt
i
m
e
d
ia i
n
f
o
r
m
a
ti
o
n
p
r
o
c
e
ss
in
g
.
K
h
a
irun
S
a
d
d
a
m
i
re
c
e
iv
e
d
th
e
B.
En
g
.
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
f
ro
m
S
y
iah
Ku
a
la
Un
iv
e
rsit
y
,
Ba
n
d
a
A
c
e
h
,
In
d
o
n
e
sia
,
in
2
0
1
5
.
He
is
c
u
rre
n
tl
y
a
P
h
D
c
a
n
d
i
d
a
te
in
P
o
stg
ra
d
u
a
te
p
ro
g
ra
m
in
En
g
in
e
e
rin
g
a
t
S
y
iah
Ku
a
la
Un
iv
e
rsity
.
He
is
a
lso
re
s
e
a
rc
h
a
ss
istan
t
in
M
u
l
ti
m
e
d
ia
a
n
d
S
ig
n
a
l
P
r
o
c
e
ss
in
g
re
se
a
rc
h
g
ro
u
p
(M
u
sig
),
E
lec
tri
c
a
l
a
n
d
Co
m
p
u
ter
En
g
i
n
e
e
rin
g
De
p
a
rtme
n
t,
S
y
i
a
h
Ku
a
la
Un
iv
e
r
sity
.
His
re
se
a
r
c
h
in
tere
sts
in
c
lu
d
e
c
o
m
p
u
ter
v
isio
n
a
n
d
ima
g
e
p
ro
c
e
ss
in
g
.
K
h
a
iru
l
M
u
n
a
d
i
re
c
e
i
v
e
d
th
e
B.
E.
d
e
g
re
e
f
ro
m
S
e
p
u
lu
h
N
o
p
e
m
b
e
r
In
stit
u
te
o
f
T
e
c
h
n
o
l
o
g
y
,
S
u
ra
b
a
y
a
,
In
d
o
n
e
sia
,
i
n
1
9
9
6
,
a
n
d
t
h
e
M
.
E
n
g
.
a
n
d
P
h
.
D.
d
e
g
re
e
s
f
ro
m
T
o
k
y
o
M
e
tro
p
o
li
tan
Un
iv
e
rsit
y
(
T
M
U),
Ja
p
a
n
,
in
2
0
0
4
a
n
d
2
0
0
7
re
sp
e
c
ti
v
e
l
y
,
a
ll
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
.
F
ro
m
1
9
9
6
t
o
1
9
9
9
,
h
e
w
a
s
w
it
h
A
lca
t
e
l
In
d
o
n
e
sia
a
s
a
s
y
ste
m
e
n
g
in
e
e
r.
S
in
c
e
1
9
9
9
,
h
e
jo
in
e
d
S
y
iah
Ku
a
la Un
iv
e
rsit
y
,
Ba
n
d
a
A
c
e
h
,
In
d
o
n
e
sia
,
a
s a
lec
tu
re
r
a
t
th
e
El
e
c
tr
ica
l
En
g
in
e
e
rin
g
De
p
a
rt
m
e
n
t.
He
w
a
s
a
v
isit
in
g
re
s
e
a
rc
h
e
r
a
t
th
e
In
f
o
rm
a
ti
o
n
a
n
d
Co
m
m
u
n
ica
ti
o
n
S
y
ste
m
s
En
g
in
e
e
rin
g
,
F
a
c
u
lt
y
o
f
S
y
ste
m
D
e
sig
n
,
T
M
U,
Ja
p
a
n
,
f
ro
m
M
a
rc
h
2
0
0
7
t
o
M
a
rc
h
2
0
0
8
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
m
u
lt
ime
d
ia sig
n
a
l
p
ro
c
e
ss
in
g
a
n
d
c
o
m
m
u
n
ica
ti
o
n
s.
Evaluation Warning : The document was created with Spire.PDF for Python.