I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
.
14
, N
o.
4
,
A
ugus
t
2025
, pp.
3412
~
3420
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
14
.i
4
.pp
3412
-
3420
3412
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
T
r
an
sf
or
m
i
n
g i
m
age
s i
n
t
o w
or
d
s:
op
t
i
c
al
c
h
ar
ac
t
e
r
r
e
c
ogn
i
t
i
on
sol
u
t
i
on
s f
or
i
m
age
t
e
xt
e
xt
r
a
c
t
i
on
Jyot
i
Wad
m
ar
e
1
, S
u
n
it
a
R
avi
n
d
r
a
P
a
t
il
2
, D
ak
s
h
it
a K
o
lt
e
1
,
K
ap
il
B
h
at
ia
1
, P
al
a
k
D
e
s
ai
1
,
G
an
e
s
h
Wad
m
ar
e
3
1
D
e
pa
r
t
m
e
nt
of
C
om
put
e
r
E
ngi
ne
e
r
i
ng, K
.J
.
S
om
a
i
ya
I
ns
t
i
t
ut
e
of
T
e
c
hnol
ogy,
M
um
ba
i
, I
ndi
a
2
S
V
K
M
'
s
N
a
r
s
e
e
M
onj
e
e
I
ns
t
i
t
ut
e
of
M
a
n
a
ge
m
e
nt
S
t
udi
e
s
, D
hul
e
, I
ndi
a
3
D
e
pa
r
t
m
e
nt
of
A
r
t
i
f
i
c
i
a
l
I
nt
e
l
l
i
ge
nc
e
a
nd D
a
t
a
S
c
i
e
nc
e
, K
.J
. S
om
a
i
ya
I
ns
t
i
t
ut
e
of
T
e
c
hnol
ogy, M
um
ba
i
, I
ndi
a
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
M
a
r
5, 2024
R
e
vi
s
e
d
F
e
b 13, 2025
A
c
c
e
pt
e
d
M
a
r
15
, 2025
Optical character recognition
(OCR)
tool is a boon
and grea
test advanc
ement
in
today’s
emerging
technology
which
has
proven
its
remarkability
in
recent
years
by
making
it
easier
for
humans
to
convert
the
textual
informa
tion
in
images
or
physical
documents
into
text
data
making
it
useful
for
an
alysis,
automati
on
processes
and
improvi
sed
producti
vity
for
different
purpose
s.
This
paper
presents
the
designing,
development
and
implementation
of
a
novel
OCR
t
ool
aiming
at
text
extraction
and
recognition
tasks.
Th
e
tool
incorpora
tes
advance
d
techniq
ues
such
as
computer
vision
and
natural
language
processing
(NLP)
which
offe
r
powerful
perform
ance
for
various
document
types.
The
performance
of
the
tool
is
subject
to
metri
cs
like
analysis
,
accuracy,
speed,
and
document
format
compatibility.
The
developed
OCR
tool
provides
an
accuracy
of
98.8%
upon
execution
p
rovi
ding
a
character
error
rate
of
2.4%
and
word
error
rate
(WER)
of
2.8%.
OC
R
tool
finds
its
applications
in
document
digitization,
personal
identifi
cation,
archival
of
valuable
document
s,
processin
g
of
invoices
,
and
other
docu
ments.
OCR
tool
holds
an
immense
amount
of
value
for
researcher
s,
practition
ers
and
many organ
izations whic
h
seek ef
fective
technique
s for
relevan
t and
a
ccurate
text extra
ction and r
ecognition ta
sks.
K
e
y
w
o
r
d
s
:
N
a
m
e
d e
nt
it
y r
e
c
ogni
ti
on
N
a
tu
r
a
l
la
ngua
ge
pr
oc
e
s
s
in
g
O
pt
ic
a
l
c
ha
r
a
c
te
r
r
e
c
ogni
ti
on
T
e
xt
e
xt
r
a
c
ti
on
T
e
xt
r
e
c
ogni
ti
on
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
J
yot
i
W
a
dm
a
r
e
D
e
pa
r
tm
e
nt
of
C
om
put
e
r
E
ngi
ne
e
r
in
g, K
. J
. S
om
a
iy
a
I
ns
ti
tu
te
of
T
e
c
hnol
ogy
S
om
a
iy
a
A
yur
vi
ha
r
C
om
pl
e
x, E
a
s
te
r
n E
xpr
e
s
s
H
ig
hw
a
y, S
io
n (
E
a
s
t)
, M
um
ba
i
400 022
, I
ndi
a
E
m
a
il
:
jy
ot
i@s
om
a
iy
a
.e
du
1.
I
N
T
R
O
D
U
C
T
I
O
N
C
om
put
e
r
vi
s
io
n,
a
s
ubf
ie
ld
of
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
,
u
s
e
s
m
a
c
hi
ne
le
a
r
ni
ng
a
nd
ne
ur
a
l
ne
twor
k
s
to
tr
a
in
s
ys
te
m
s
how
to
e
xt
r
a
c
t
us
e
f
ul
in
f
or
m
a
ti
on
f
r
om
di
gi
ta
l
in
pu
ts
s
uc
h
a
s
im
a
ge
s
a
nd
vi
de
o
s
.
T
he
s
e
a
lg
or
it
hm
s
c
a
n
th
e
n
s
ugge
s
t
a
c
ti
vi
ti
e
s
or
id
e
nt
if
y
pr
o
bl
e
m
s
in
vi
s
ua
l
da
ta
.
S
im
il
a
r
ly
,
to
e
na
bl
e
c
om
put
e
r
s
c
om
pr
e
he
nd,
in
te
r
pr
e
t
a
nd
pr
oduc
e
m
e
a
ni
ngf
ul
hum
a
n
la
ngua
ge
,
na
tu
r
a
l
l
a
ngua
ge
pr
oc
e
s
s
in
g
(
N
L
P
)
is
u
s
e
d
[
1]
.
N
L
P
te
c
hni
que
s
in
c
lu
de
t
e
xt
unde
r
s
ta
ndi
ng,
tr
a
ns
la
ti
on,
s
e
nt
im
e
nt
a
na
ly
s
is
,
s
pe
e
c
h
r
e
c
ogni
ti
on,
te
xt
ge
ne
r
a
ti
on,
in
f
or
m
a
ti
on
r
e
t
r
ie
va
l,
na
m
e
d
e
nt
it
y
r
e
c
ogni
ti
on
(
N
E
R
)
,
a
nd
qu
e
s
ti
on
a
ns
w
e
r
in
g.
A
ppl
ic
a
ti
ons
in
c
lu
de
vi
r
tu
a
l
a
s
s
is
t
a
nt
s
, t
r
a
ns
la
ti
on s
e
r
vi
c
e
s
,
s
e
nt
im
e
nt
a
na
ly
s
i
s
t
ool
s
, a
nd i
nf
or
m
a
ti
on r
e
tr
ie
va
l
s
ys
te
m
s
.
O
pt
ic
a
l
c
ha
r
a
c
te
r
r
e
c
ogni
ti
on
(
O
C
R
)
i
s
a
c
om
bi
na
ti
on
of
c
o
m
put
e
r
vi
s
io
n
a
nd
N
L
P
th
a
t
c
onve
r
ts
pr
in
te
d
or
ha
ndw
r
it
te
n
te
xt
f
r
om
phot
os
in
to
e
di
ta
bl
e
,
m
a
c
hi
ne
-
r
e
a
da
bl
e
te
xt
.
O
C
R
e
xt
r
a
c
ts
te
xt
f
r
om
im
a
ge
s
or
doc
um
e
nt
s
by
lo
c
a
ti
ng
te
xt
a
r
e
a
s
,
s
e
pa
r
a
ti
ng
te
xt
,
id
e
nt
i
f
yi
ng c
ha
r
a
c
te
r
s
,
a
nd
out
put
ti
ng
th
e
r
e
c
ogni
z
e
d
te
xt
.
I
t
di
gi
ti
z
e
s
doc
um
e
nt
s
a
nd
e
na
bl
e
s
te
xt
s
e
a
r
c
he
s
in
im
a
ge
s
.
T
h
e
pr
opos
e
d
O
C
R
s
y
s
te
m
w
a
s
c
r
e
a
t
e
d
to
c
ol
le
c
t
phot
ogr
a
phs
,
e
xt
r
a
c
t
te
xt
us
in
g
T
e
s
s
e
r
a
c
t
O
C
R
us
in
g
P
yt
e
s
s
e
r
a
c
t,
a
nd
c
le
a
n
th
e
m
.
I
n
or
de
r
f
or
s
pa
C
y
to
tr
a
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
T
r
ans
fo
r
m
in
g i
m
age
s
i
nt
o w
or
d
s
:
opt
ic
al
c
har
ac
te
r
r
e
c
ogni
ti
on s
ol
ut
io
ns
f
or
…
(
J
y
ot
i
W
adm
ar
e
)
3413
th
e
m
ode
l,
N
E
R
da
ta
is
m
a
nua
ll
y
la
be
le
d
us
in
g
be
gi
nni
ng,
i
ns
id
e
,
out
s
id
e
(
B
I
O
)
ta
ggi
ng
a
nd
or
ga
ni
z
e
d.
T
he
m
ode
l
pr
e
di
c
ts
e
nt
it
ie
s
,
w
hi
c
h
a
r
e
r
e
pr
e
s
e
nt
e
d
in
di
s
pl
a
C
y
a
nd
hi
ghl
ig
ht
e
d
in
im
a
ge
s
.
T
he
s
e
ta
s
ks
a
r
e
s
im
pl
if
ie
d
by
w
e
b
s
of
twa
r
e
,
w
hi
c
h
s
a
ve
s
e
xt
r
a
c
te
d
te
xt
in
a
n
E
xc
e
l
f
il
e
.
T
he
pr
im
a
r
y
a
lg
or
it
h
m
s
us
e
d
a
r
e
C
a
nny
e
dge
de
te
c
ti
on
a
nd D
ougl
a
s
-
P
e
uc
ke
r
f
or
e
dge
de
te
c
ti
on a
nd pol
yl
in
e
s
im
pl
if
ic
a
ti
on.
O
C
R
ha
s
a
w
id
e
r
a
nge
of
us
e
s
,
in
c
lu
di
ng
di
gi
ti
z
in
g
s
tu
de
nt
r
e
c
or
ds
,
m
e
di
c
a
l
doc
um
e
nt
s
,
a
nd
in
voi
c
in
g
in
e
duc
a
ti
on,
he
a
lt
hc
a
r
e
,
a
nd
f
in
a
nc
e
,
a
s
w
e
ll
a
s
be
n
e
f
it
in
g
ot
he
r
in
dus
tr
ie
s
by
im
pr
ovi
ng
di
gi
ta
li
z
a
ti
on
a
nd
da
ta
a
c
c
e
s
s
ib
il
it
y
[
2]
.
T
he
c
onve
r
s
io
n
of
ha
ndw
r
it
te
n
a
nd
pr
in
te
d
pa
pe
r
s
in
to
e
di
ta
bl
e
,
s
e
a
r
c
ha
bl
e
da
t
a
s
im
pl
if
ie
s
a
va
r
ie
ty
of
hum
a
n
ope
r
a
ti
ons
by
s
pe
e
di
ng
da
ta
e
nt
r
y,
r
e
c
or
d
-
ke
e
pi
ng,
a
nd
in
f
or
m
a
ti
on
r
e
tr
ie
va
l.
B
y
im
pr
ovi
ng
a
c
c
ur
a
c
y
a
nd
e
f
f
ic
ie
nc
y,
th
is
c
ha
nge
lo
w
e
r
s
th
e
pos
s
i
bi
li
ty
of
e
r
r
or
s
a
nd
f
a
c
il
it
a
te
s
f
a
s
te
r
a
c
c
e
s
s
to
e
s
s
e
nt
ia
l
doc
um
e
nt
s
.
I
t
a
l
s
o
im
pr
ove
s
da
ta
or
ga
ni
z
a
ti
on
a
nd
a
dm
in
is
tr
a
ti
on,
m
a
ki
ng
it
e
a
s
ie
r
to
e
va
lu
a
te
a
nd
di
s
tr
ib
ut
e
in
f
or
m
a
ti
on
a
c
r
os
s
m
ul
ti
pl
e
pl
a
tf
or
m
s
a
nd
s
ys
te
m
s
[
3]
.
W
hi
le
O
C
R
w
or
ks
w
e
ll
w
it
h
m
ode
r
n
te
xt
,
it
is
le
s
s
a
c
c
ur
a
te
w
it
h
hi
s
to
r
ic
doc
um
e
nt
s
a
nd
out
da
te
d
di
gi
ta
li
z
a
ti
on
t
e
c
hnol
ogi
e
s
.
A
s
a
r
e
s
ul
t,
noi
s
y
di
gi
ti
z
e
d
te
xt
s
r
e
qui
r
e
pos
t
-
c
or
r
e
c
ti
on
to
im
pr
ove
O
C
R
r
e
s
ul
ts
,
w
hi
c
h
a
r
e
c
r
it
ic
a
l
f
or
in
f
or
m
a
ti
on
r
e
tr
ie
va
l
a
nd
N
L
P
a
ppl
ic
a
ti
ons
[
4]
.
T
he
s
tr
uc
tu
r
e
of
th
is
doc
um
e
nt
is
a
s
f
ol
lo
w
s
:
s
e
c
ti
on
1
is
th
e
in
tr
oduc
ti
on
to
w
ha
t
is
O
C
R
.
S
e
c
ti
on
2
w
e
s
tu
dy
a
bout
th
e
va
r
io
us
to
ol
s
f
or
O
C
R
a
lr
e
a
dy
a
va
il
a
bl
e
a
n
d
th
e
ir
li
te
r
a
tu
r
e
s
ur
ve
y,
s
e
c
ti
on
3
pr
ovi
de
s
a
n
ove
r
vi
e
w
of
th
e
s
ys
te
m
w
hi
c
h
is
de
ve
lo
pe
d,
s
e
c
ti
on
4
di
s
c
us
s
e
s
th
e
r
e
s
ul
ts
of
th
e
to
ol
m
a
de
.
S
e
c
ti
on
5
p
r
e
s
e
nt
s
th
e
c
onc
lu
s
io
ns
.
2.
R
E
L
A
T
E
D
WORK
T
he
s
tr
ong
f
ounda
ti
on
to
de
ve
lo
p
O
C
R
to
ol
c
a
m
e
th
r
ough
a
t
hor
ough
s
ur
ve
y
of
r
e
s
e
a
r
c
h
pa
p
e
r
s
.
I
t
he
lp
e
d
in
unde
r
s
ta
ndi
ng
th
e
ne
e
d
f
or
th
e
de
ve
lo
pm
e
nt
of
th
e
O
C
R
to
ol
s
f
or
te
xt
r
e
c
ogni
ti
on
f
r
om
th
e
im
a
ge
s
.
T
he
t
e
xt
obt
a
in
e
d f
r
om
t
he
e
xt
r
a
c
ti
on pr
oc
e
s
s
i
s
s
ub
s
e
que
nt
ly
s
t
or
e
d w
it
hi
n a
n E
xc
e
l
f
il
e
.
O
la
da
yo
[
5]
f
oc
us
e
s
on
u
s
in
g
O
C
R
te
c
hnol
ogy
to
c
onve
r
t
a
nd
k
e
e
p
m
a
ny
pa
pe
r
s
in
hi
s
to
r
ic
a
l
a
r
c
hi
ve
s
di
gi
ta
ll
y. R
e
gul
a
r
s
c
a
nne
r
s
s
c
a
n i
m
a
ge
s
f
r
om
t
he
doc
um
e
nt
s
s
u
c
h t
ha
t
th
e
y c
a
nnot
be
us
e
d on s
c
r
e
e
n or
e
di
te
d
w
it
h
a
ny
s
of
twa
r
e
us
e
d
f
or
a
ny
ot
he
r
doc
um
e
nt
ty
pe
.
T
hi
s
s
tu
dy
unve
il
s
a
n
O
C
R
s
o
f
twa
r
e
w
hi
c
h
is
a
bl
e
to
c
onve
r
t
of
f
li
ne
ty
pe
d
a
nd
ha
ndw
r
it
te
n
doc
um
e
nt
s
in
to
te
xt
f
or
m
s
th
a
t
c
a
n
be
e
di
te
d.
B
y
ut
il
iz
in
g
a
m
or
phol
ogi
c
a
l
c
or
r
e
la
ti
on t
e
c
hni
que
, t
hi
s
s
ys
te
m
e
nh
a
nc
e
s
t
h
e
e
f
f
ic
ie
nc
y of
t
e
xt
m
a
ppi
ng a
nd r
e
c
ogni
t
io
n
[
5]
.
A
dj
e
te
y
a
nd
M
a
nu
[
6]
pr
e
s
e
nt
a
nove
l
te
c
hni
que
to
e
nha
nc
e
i
m
a
ge
r
e
tr
ie
va
l
s
y
s
te
m
s
(
I
R
S
s
)
.
T
he
ir
a
ppr
oa
c
h
in
te
gr
a
te
s
a
T
e
s
s
e
r
a
c
t
O
C
R
e
ngi
ne
,
a
nd
a
n
e
nha
n
c
e
d
te
xt
-
m
a
tc
hi
ng
a
lg
or
it
hm
,
le
ve
r
a
gi
ng
th
e
le
ve
ns
ht
e
in
a
lg
or
it
hm
. E
xpe
r
im
e
nt
a
l
r
e
s
ul
ts
de
m
ons
tr
a
te
a
100%
s
uc
c
e
s
s
r
a
te
i
n r
e
tr
ie
vi
ng t
he
a
ppr
opr
ia
te
f
il
e
ba
s
e
d
on
pa
r
ti
a
l
que
r
y
im
a
ge
s
,
s
how
c
a
s
in
g
th
e
e
f
f
e
c
ti
ve
ne
s
s
of
th
is
in
te
gr
a
te
d
a
ppr
oa
c
h
in
i
m
pr
ovi
ng
i
m
a
ge
r
e
tr
ie
va
l
a
c
c
ur
a
c
y
[
6]
.
Z
hu
e
t
al
[
7
]
in
tr
oduc
e
S
hot
V
is
,
a
nove
l
a
ppr
oa
c
h
to
c
a
pt
ur
e
t
e
xt
im
a
ge
s
f
r
om
m
obi
le
de
vi
c
e
s
a
nd
pr
oc
e
s
s
th
e
s
e
te
xt
im
a
ge
s
to
s
to
r
e
c
ha
r
a
c
te
r
s
a
s
s
tr
uc
tu
r
e
d
da
ta
.
I
t
a
ll
ow
s
us
e
r
s
to
li
nk
vi
s
ua
l
f
or
m
s
to
th
e
unde
r
ly
in
g
da
ta
a
nd
ge
ne
r
a
te
vi
s
ua
li
z
a
ti
ons
th
r
ough
to
uc
h
-
ba
s
e
d
in
te
r
a
c
ti
ons
.
W
it
h
a
s
im
pl
e
c
li
c
k
of
th
e
c
a
m
e
r
a
,
S
hot
V
is
s
w
if
tl
y
s
um
m
a
r
iz
e
s
te
xt
f
r
om
im
a
ge
s
in
to
w
or
d
c
lo
uds
,
s
c
a
tt
e
r
pl
ot
s
,
a
nd
va
r
io
us
ot
he
r
vi
s
ua
li
z
a
ti
ons
, e
n
a
bl
in
g i
nt
e
r
a
c
ti
ve
e
xpl
or
a
ti
on of
t
e
xt
da
ta
c
a
pt
ur
e
d vi
a
s
m
a
r
tp
hone
c
a
m
e
r
a
s
[
7]
.
S
uddul
a
nd
S
e
gui
n
[
8]
r
e
c
om
m
e
nd
a
pr
oc
e
s
s
of
c
us
to
m
e
r
r
e
gi
s
tr
a
ti
on
us
in
g
de
e
p
le
a
r
ni
ng
-
ba
s
e
d
O
C
R
te
c
hnol
ogy
w
hi
c
h
c
a
n
be
ut
il
iz
e
d
f
or
a
ut
om
a
ti
c
te
xt
e
xt
r
a
c
ti
on
f
r
om
im
a
ge
s
of
I
D
c
a
r
ds
.
T
he
f
ir
s
t
s
te
p
in
vol
ve
s
th
e
te
xt
s
pot
s
id
e
nt
if
ic
a
ti
on
by
a
pr
opr
ie
ta
r
y
U
-
N
e
t
im
a
ge
s
e
gm
e
nt
a
ti
on
a
lg
or
it
hm
a
nd
th
e
ot
he
r
s
te
p
is
to
r
e
c
ogni
z
e
c
ha
r
a
c
te
r
s
a
nd
f
or
m
a
ti
on
of
w
or
ds
in
c
onvolut
io
na
l
r
e
c
ur
r
e
nt
ne
ur
a
l
ne
two
r
k
(
C
R
N
N
)
w
it
h
lo
ng
s
hor
t
-
te
r
m
m
e
m
or
y
(
L
S
T
M
)
c
e
ll
s
.
T
he
e
xpe
r
im
e
nt
w
a
s
c
a
r
r
ie
d out
on
M
a
ur
it
iu
s
’
na
ti
ona
l
id
e
nt
it
y
c
a
r
d
a
nd
it
yi
e
ld
s
0.70
in
te
r
s
e
c
ti
on ove
r
uni
on (
I
oU
)
s
c
or
e
a
nd 98%
pi
xe
l
a
c
c
ur
a
c
y
[
8]
.
S
a
ti
r
a
pi
w
ong
a
nd
S
ir
ib
or
vor
nr
a
ta
na
kul
[
9]
a
ddr
e
s
s
e
d
th
e
c
ha
ll
e
nge
s
of
pr
oc
e
s
s
in
g
T
ha
i
in
voi
c
e
s
f
or
bus
in
e
s
s
p
a
ym
e
nt
s
,
w
hi
c
h
tr
a
di
ti
ona
ll
y
r
e
qui
r
e
s
e
xt
e
ns
iv
e
m
a
n
ua
l
e
f
f
or
ts
a
nd
te
m
pl
a
te
m
a
tc
hi
ng.
T
he
pa
pe
r
in
tr
oduc
e
s
bi
di
r
e
c
ti
ona
l
lo
ng
s
hor
t
-
te
r
m
m
e
m
or
y
-
c
ondi
ti
ona
l
r
a
ndom
f
ie
ld
(
B
iL
S
T
M
-
CRF
)
de
e
p
le
a
r
ni
ng
m
ode
l
w
hi
c
h
us
e
s
m
ix
in
g
of
w
or
ds
a
nd
c
ha
r
a
c
te
r
s
s
pe
c
if
ic
a
l
ly
f
or
T
ha
i
in
voi
c
e
s
.
T
hi
s
m
ode
l
hi
ghl
ig
ht
e
d
a
c
c
ur
a
te
F
1
-
s
c
or
e
m
e
tr
ic
s
, pr
e
c
i
s
io
n s
c
or
e
a
nd r
e
c
a
ll
. T
he
qu
a
li
ty
of
O
C
R
w
a
s
hi
ghl
ig
ht
e
d us
in
g F
1
-
s
c
or
e
[
9]
.
L
e
e
[
10]
de
f
in
e
s
t
he
us
e
of
m
a
ny dif
f
e
r
e
nt
l
ib
r
a
r
ie
s
, s
pe
c
if
ic
a
ll
y
us
in
g i
nt
e
r
li
br
a
r
y l
oa
n (
I
L
L
)
. T
o t
e
s
t
th
e
pr
oc
e
s
s
, 20 c
opi
e
s
of
a
r
ti
c
le
s
w
e
r
e
s
e
nt
t
o t
e
s
t
th
e
a
c
c
ur
a
c
y
of
A
dobe
A
c
r
oba
t
P
r
o D
C
t
o c
r
e
a
te
s
e
a
r
c
ha
bl
e
P
D
F
s
. T
he
a
c
c
ur
a
c
y of
a
ut
om
a
te
d O
C
R
r
e
s
ul
t
s
w
a
s
c
a
lc
ul
a
te
d a
nd ma
nua
l
c
or
r
e
c
ti
ons
w
e
r
e
m
a
de
a
f
te
r
t
ha
t
to
a
voi
d
pr
obl
e
m
s
w
hi
c
h
w
oul
d
pr
ovi
de
a
good
in
it
ia
ti
ve
f
or
I
L
L
to
pr
ovi
de
pa
tr
ons
w
it
h
m
a
te
r
ia
ls
th
a
t
a
r
e
a
c
c
e
s
s
ib
le
[
10]
.
M
a
ni
va
nna
n
e
t
al
.
[
11]
p
r
opos
e
d
a
n
e
ne
r
gy
-
e
f
f
ic
ie
nt
I
o
T
m
ode
l
f
or
pr
e
di
c
ti
ng
ha
ndw
r
it
te
n
pr
e
s
c
r
ip
ti
on
of
doc
to
r
s
.
I
t
c
ons
id
e
r
s
m
a
ki
ng
us
e
of
a
tr
ib
oe
le
c
tr
ic
s
m
a
r
t
r
e
c
ogni
ti
on
s
ys
te
m
f
or
r
e
c
ogni
ti
on
o
f
m
e
di
c
a
l
te
r
m
s
a
nd
is
c
on
s
id
e
r
e
d
to
be
r
obus
t.
T
h
e
s
ys
t
e
m
r
e
s
ul
ts
in
di
gi
ta
l
twi
n
de
ve
lo
pm
e
nt
f
or
m
oni
to
r
in
g
s
ys
te
m
s
to
tr
a
c
k
us
a
ge
w
he
r
e
in
di
vi
dua
l
pr
e
s
c
r
ip
ti
ons
c
a
n
b
e
de
v
e
lo
pe
d
a
nd
a
na
ly
z
e
d.
T
he
r
e
tu
r
n
on
in
ve
s
tm
e
nt
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 4,
A
ugus
t
2025
:
3412
-
3420
3414
(
R
O
I
)
of
th
e
s
y
s
te
m
is
e
va
lu
a
te
d
on
th
e
b
a
s
is
of
di
f
f
e
r
e
nt
pa
r
a
m
e
te
r
s
s
uc
h
a
s
O
C
R
,
a
c
c
ur
a
c
y,
s
e
n
s
it
iv
it
y,
s
pe
c
if
ic
it
y
,
a
nd s
e
n
s
it
iv
it
y a
na
ly
s
is
[
11]
.
P
oodi
kka
la
m
a
nd
L
oga
na
th
a
n
[
12]
f
oc
us
e
d
on
c
ogni
ti
ve
pr
oc
e
s
s
in
g
f
or
O
C
R
.
T
he
a
ut
hor
s
id
e
nt
if
ie
d
s
c
a
le
-
in
va
r
ia
nt
tr
a
ns
f
or
m
in
g
f
e
a
tu
r
e
(
S
I
F
T
)
de
s
c
r
ip
to
r
s
w
hi
c
h
us
e
s
two
f
unc
ti
ons
.
T
he
ot
he
r
pr
oc
e
s
s
,
na
m
e
ly
R
oot
S
I
F
T
gi
ve
s
e
xc
e
pt
io
na
l
r
e
s
ul
ts
w
it
hout
s
to
r
a
ge
r
e
qui
r
e
m
e
nt
s
or
c
om
put
a
ti
ona
l
c
om
pl
e
xi
ty
.
A
r
ti
f
ic
ia
l
be
e
c
ol
ony
(
A
B
C
)
is
u
s
e
d
f
or
id
e
nt
if
ic
a
ti
on
of
E
ngl
is
h
la
ngua
ge
c
ha
r
a
c
te
r
s
.
T
h
e
a
c
c
ur
a
c
y
of
num
be
r
s
,
a
lp
ha
num
e
r
ic
c
ha
r
a
c
te
r
s
in
c
lu
di
ng
s
m
a
ll
a
nd
bi
g
le
tt
e
r
s
is
te
s
te
d
a
nd
it
is
f
ound
th
a
t
A
B
C
a
lg
or
it
hm
ha
s
m
a
xi
m
um
e
f
f
ic
ie
nc
y of
a
r
ound 97.3077%
[
12]
.
M
ohd
e
t
al
.
[
13]
f
oc
us
s
e
s
on
Q
ur
a
ni
c
O
C
R
w
hi
c
h
i
s
de
ve
lo
pe
d
us
in
g
c
onvolut
io
na
l
ne
ur
a
l
ne
twor
k
(
C
N
N
)
a
nd
r
e
c
ur
r
e
nt
ne
ur
a
l
ne
twor
k
(
R
N
N
)
.
B
a
s
e
d
on
th
e
pr
i
nt
e
d
ve
r
s
io
n
of
th
e
H
ol
y
Q
ur
a
n,
a
ne
w
da
ta
s
e
t
w
a
s
de
ve
lo
pe
d
w
hi
c
h
r
e
c
ogni
z
e
s
th
e
Q
ur
a
ni
c
im
a
g
e
’
s
di
a
c
r
it
i
c
te
xt
.
T
w
o
m
ode
l
s
c
om
pa
r
e
d
f
or
A
r
a
bi
c
te
xt
r
e
c
ogni
ti
on
w
e
r
e
L
S
T
M
a
nd
ga
t
e
d
r
e
c
ur
r
e
nt
uni
t
(
G
R
U
)
w
he
r
e
a
publ
ic
da
ta
ba
s
e
w
a
s
bui
lt
a
nd
it
a
c
hi
e
ve
d
a
c
c
ur
a
c
y
of
98%
w
it
h
va
li
da
ti
on
d
a
ta
a
nd
95%
w
or
d
r
e
c
ogni
ti
on
r
a
te
(
W
R
R
)
a
nd
c
ha
r
a
c
te
r
r
e
c
ogni
ti
on
r
a
te
(
CRR
)
of
99%
i
n t
he
t
e
s
t
da
ta
s
e
t
[
13]
.
H
a
s
s
a
n
e
t
al
.
[
14]
hi
ghl
ig
ht
s
th
a
t
A
r
a
bi
c
s
c
e
ne
te
xt
r
e
c
ogni
ti
o
n
is
a
c
om
pl
e
x
pa
r
t
in
unde
r
s
ta
ndi
ng
s
c
e
ne
s
y
s
te
m
s
. T
he
us
e
of
A
r
a
bi
c
in
vol
vi
ng
L
a
ti
n
c
ha
r
a
c
te
r
s
is
li
m
it
e
d
in
de
e
p
le
a
r
ni
ng
m
e
th
ods
.
T
he
da
ta
s
e
t
e
va
lu
a
te
s
th
r
e
e
p
a
r
a
m
e
te
r
s
-
us
e
of
de
e
p
l
e
a
r
ni
ng
te
c
hni
que
s
,
id
e
nt
if
yi
ng
c
ha
ll
e
nge
s
in
A
r
a
bi
c
te
xt
a
nd
in
ve
s
ti
ga
ti
on of
bi
li
ngua
l
m
ode
ls
. T
he
da
ta
s
e
t
us
e
d he
lp
s
i
n pr
o
vi
di
ng dir
e
c
ti
ons
f
or
f
ut
ur
e
r
e
s
e
a
r
c
h
[
14]
.
M
a
lh
ot
r
a
a
nd A
ddi
s
[
15]
hi
gh
li
ght
th
e
E
th
io
pi
c
ha
ndw
r
i
tt
e
n t
e
x
t
r
e
c
ogni
ti
on
us
in
g s
e
que
nt
ia
l
f
e
a
tu
r
e
e
xt
r
a
c
ti
on
a
nd
e
f
f
ic
ie
nt
r
e
c
ogni
ti
on
us
in
g
a
n
e
nd
-
to
-
e
nd
s
tr
a
te
gy.
T
he
a
r
c
hi
te
c
tu
r
e
of
th
e
m
ode
l
in
c
lu
de
s
a
n
a
tt
e
nt
io
n
m
e
c
ha
ni
s
m
a
nd
a
c
onne
c
ti
oni
s
t
te
m
por
a
l
c
la
s
s
if
ic
a
ti
on
ut
il
iz
in
g
s
e
ve
n
C
N
N
s
a
nd
two
R
N
N
s
a
r
e
u
s
e
d
f
or
m
ode
l
t
r
a
in
in
g. T
he
a
c
c
ur
a
c
y of
c
ha
r
a
c
te
r
e
r
r
or
r
a
te
(
C
E
R
)
obt
a
in
e
d w
a
s
17.95%
f
or
te
s
t
s
e
t
I
a
nd 29.95%
f
or
te
s
t
s
e
t
II
[
15]
.
W
a
ng
e
t
al
.
[
16]
a
tt
e
m
pt
e
d
to
im
pr
ove
C
hi
ne
s
e
O
C
R
a
c
c
ur
a
c
y
b
y
c
r
e
a
ti
ng
a
hybr
id
r
e
c
ogni
ti
on
m
ode
l
th
a
t
w
a
s
s
ui
te
d
to
th
e
la
ngua
ge
'
s
di
s
ti
nc
ti
ve
f
e
a
tu
r
e
s
.
T
hi
s
a
ppr
o
a
c
h
pr
e
-
f
il
te
r
s
im
a
ge
in
te
r
f
e
r
e
nc
e
a
nd
m
odi
f
ie
s
c
ha
r
a
c
te
r
a
s
pe
c
t
r
a
ti
o
s
pr
io
r
t
o O
C
R
pr
oc
e
s
s
in
g.
E
xpe
r
im
e
nt
s
r
e
ve
a
le
d t
ha
t
i
m
a
ge
pr
oc
e
s
s
in
g
r
a
is
e
d
T
e
s
s
e
r
a
c
t
-
O
C
R
'
s
c
or
r
e
c
t
id
e
nt
if
ic
a
ti
on r
a
te
by a
bout
12%
, w
hi
ls
t
N
L
P
in
c
r
e
a
s
e
d a
c
c
ur
a
c
y by a
bout
5%
[
16]
.
S
ha
hi
r
a
a
nd
L
ij
iy
a
[
17
]
pr
opos
e
th
a
t
f
or
th
e
e
a
s
e
of
c
om
m
uni
c
a
ti
on,
te
xt
ua
l
da
ta
is
s
uppor
te
d
w
it
h
gr
a
phi
c
a
l
r
e
pr
e
s
e
nt
a
ti
ons
but
t
hi
s
i
s
not
a
ppl
ic
a
bl
e
f
or
bl
in
d or
v
is
ua
ll
y i
m
pa
ir
e
d pe
opl
e
.
T
he
pa
p
e
r
f
oc
us
e
s
on
e
xt
r
a
c
ti
ng
va
lu
a
bl
e
in
f
or
m
a
ti
on
or
c
r
it
ic
a
l
da
ta
f
r
om
c
ha
r
ts
or
gr
a
p
hs
.
L
oc
a
li
z
a
ti
on
a
nd
c
la
s
s
if
ic
a
ti
on
a
r
e
te
c
hni
que
s
th
a
t
c
a
n
be
im
pl
e
m
e
nt
e
d
u
s
in
g
de
e
p
le
a
r
ni
ng.
T
h
e
pa
pe
r
s
ugge
s
ts
th
e
u
s
e
of
hum
a
n
c
om
put
e
r
in
te
r
a
c
ti
on
a
nd
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
te
c
hni
que
s
to
a
ut
om
a
te
e
x
tr
a
c
ti
on
of
da
ta
a
nd
pr
ovi
de
it
s
de
s
c
r
ip
ti
on
f
o
r
vi
s
ua
ll
y i
m
pa
ir
e
d s
e
c
ti
ons
of
s
o
c
ie
ty
[
17]
.
P
ol
a
nc
ic
e
t
al
.
[
18]
in
ve
s
ti
ga
te
s
th
e
tr
a
ns
f
or
m
a
ti
on
of
ha
nd
-
dr
a
w
n
di
a
gr
a
m
s
to
di
gi
ta
ll
y
d
r
a
w
n
di
a
gr
a
m
s
us
in
g
O
C
R
.
I
t
s
ugge
s
ts
s
ui
ta
bl
e
s
ol
ut
io
ns
ba
s
e
d
on
T
e
ns
or
F
lo
w
w
hi
c
h
pr
ovi
de
a
c
c
ur
a
te
r
e
s
ul
ts
f
o
r
di
f
f
e
r
e
nt
e
le
m
e
nt
s
or
s
e
c
ti
ons
of
ha
nd
-
dr
a
w
n
di
a
gr
a
m
s
a
nd
th
e
i
r
e
le
m
e
nt
s
.
I
t
m
a
k
e
s
u
s
e
of
di
f
f
e
r
e
nt
s
ta
ti
s
ti
c
a
l
a
ppr
oa
c
he
s
li
ke
B
a
ye
s
ia
n
c
la
s
s
if
ie
r
,
de
c
is
io
n
tr
e
e
c
la
s
s
if
ie
r
,
ne
ur
a
l
ne
twor
k
c
la
s
s
if
ie
r
,
ne
a
r
e
s
t
ne
ig
hbor
s
c
la
s
s
if
ie
r
, s
ynt
a
c
ti
c
a
ppr
oa
c
h
f
or
t
e
xt
r
e
c
ogni
ti
on
[
18]
.
U
e
da
e
t
al
.
[
19]
in
ve
s
ti
ga
te
s
th
e
t
e
xt
-
ba
s
e
d
im
a
ge
c
a
pt
io
ni
ng
m
e
t
hod
w
hi
c
h
is
us
e
d
to
pr
ovi
de
c
a
pt
io
ns
to
th
e
im
a
ge
s
in
th
e
f
or
m
of
te
xt
m
a
ki
ng
us
e
of
O
C
R
.
I
t
u
s
e
s
a
pr
e
-
tr
a
in
e
d
c
ont
r
a
s
ti
ve
la
ngua
ge
-
im
a
ge
pr
e
-
tr
a
in
in
g (
C
L
I
P
)
m
ode
l
to
i
m
pr
ove
a
nd e
nha
nc
e
i
m
a
ge
s
us
in
g l
in
gui
s
ti
c
f
e
a
tu
r
e
s
of
O
C
R
. I
t
a
l
s
o i
nt
r
oduc
e
s
two
ne
w
a
tt
e
nt
io
n
m
ode
ls
to
s
tr
e
ngt
he
n
th
e
tr
a
ns
f
or
m
a
ti
on
a
r
c
hi
te
c
tu
r
e
of
r
e
pr
e
s
e
nt
a
ti
on
of
im
a
ge
s
in
w
hi
c
h
th
e
pr
opos
e
d s
ys
te
m
out
p
e
r
f
or
m
s
t
he
T
e
xt
C
a
p
s
da
ta
s
e
t
[
19]
.
W
u
e
t
al
.
[
20]
pr
opos
e
s
a
two
-
le
ve
l
r
e
c
ti
f
ic
a
ti
on
a
tt
e
nt
io
n
ne
t
w
or
k
(
T
R
A
N
)
to
r
e
c
ti
f
y
a
nd
id
e
nt
if
y
te
xt
s
.
I
t
c
ons
is
ts
of
two
le
ve
ls
-
f
ir
s
t
is
two
-
le
ve
l
r
e
c
ti
f
ic
a
ti
on
ne
twor
k
(
T
O
R
N
)
w
hi
c
h
is
us
e
d
to
r
e
s
ol
ve
ge
om
e
tr
ic
a
l
c
ons
tr
a
in
ts
u
s
in
g
pi
xe
l
le
ve
l
a
dj
u
s
tm
e
nt
a
nd
gi
ve
c
le
a
r
te
xt
a
nd
s
e
c
ond
i
s
a
tt
e
nt
io
n
-
ba
s
e
d
r
e
c
ogni
ti
on ne
twor
k (
A
B
R
N
)
w
hi
c
h i
s
us
e
d t
o r
e
c
ogni
z
e
t
e
xt
i
n
r
e
c
ti
f
ie
d i
m
a
ge
s
.
T
o ha
ndl
e
ot
he
r
va
r
ia
ti
ons
,
a
ne
w
c
ha
nne
l
a
nd
ke
r
ne
l
w
is
e
a
tt
e
nt
io
n
uni
t
is
de
ve
lo
pe
d.
T
he
s
ta
te
-
of
-
a
r
t
pe
r
f
or
m
a
nc
e
is
a
c
hi
e
ve
d
a
s
a
r
e
s
ul
t
of
t
hi
s
e
xpe
r
im
e
nt
a
ti
on c
onduc
te
d
[
20]
.
Z
ha
ng
e
t
al
.
[
21]
f
oc
us
e
d
-
on c
ha
ll
e
nge
s
f
a
c
e
d due
t
o t
e
xt
r
e
a
di
n
g of
di
f
f
e
r
e
nt
te
xt
i
m
a
ge
s
. S
e
que
nc
e
-
li
ke
im
a
ge
s
a
r
e
di
f
f
ic
ul
t
to
pr
e
di
c
t
a
nd
c
onve
nt
io
na
l
m
e
th
od
s
d
o
not
a
li
gn
th
e
m
a
s
c
ha
r
a
c
te
r
in
f
or
m
a
ti
on.
T
h
e
m
e
th
od
us
e
d
to
a
li
gn
s
e
que
nt
ia
l
im
a
g
e
s
i
s
nove
l
a
dve
r
s
a
r
ia
l
s
e
qu
e
nc
e
-
to
-
s
e
que
nc
e
dom
a
in
a
d
a
pt
a
ti
on
(
A
S
S
D
A
)
w
hi
c
h
m
in
e
s
lo
c
a
l
r
e
gi
on
s
c
ont
a
in
in
g
c
h
a
r
a
c
te
r
s
a
nd
a
li
gns
th
e
m
in
a
n
a
dve
r
s
a
r
ia
l
m
a
nn
e
r
.
A
f
te
r
pe
r
f
or
m
in
g
e
xt
e
ns
iv
e
t
e
xt
r
e
c
ogni
ti
on, i
t
is
pr
ove
d t
ha
t
A
S
S
D
A
i
s
e
f
f
ic
ie
nt
to
t
r
a
ns
f
e
r
s
e
que
nc
e
k
now
le
dg
e
[
21]
.
Y
ıl
dı
z
[
22]
put
s
f
or
w
a
r
d a
nove
l
t
e
c
hni
que
t
o
e
m
pl
oy c
or
r
e
c
ti
on
of
gr
a
m
m
a
ti
c
a
l
e
r
r
or
s
of
te
n f
ound in
O
C
R
w
hi
c
h
in
vol
ve
s
c
or
r
e
c
ti
ng
s
ynt
a
x
a
s
w
e
ll
a
s
s
e
m
a
nt
ic
s
by
c
ons
id
e
r
in
g
how
of
te
n
s
pe
c
if
ic
c
om
bi
na
ti
on
s
of
w
or
ds
oc
c
ur
in
s
e
nt
e
nc
e
s
a
lo
ng
s
id
e
a
ppl
yi
ng
r
e
c
ur
s
io
n.
I
t
c
o
m
put
e
s
f
r
e
que
nc
y
f
or
e
ve
r
y
pa
ir
of
w
or
ds
th
a
t
oc
c
ur
one
a
f
te
r
a
not
he
r
w
it
hi
n
a
ny
gi
ve
n
body
of
te
xt
s
be
f
or
e
s
e
tt
in
g
up
a
c
or
r
e
c
ti
ona
l
hub
w
hi
c
h
c
ons
i
s
ts
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
T
r
ans
fo
r
m
in
g i
m
age
s
i
nt
o w
or
d
s
:
opt
ic
al
c
har
ac
te
r
r
e
c
ogni
ti
on s
ol
ut
io
ns
f
or
…
(
J
y
ot
i
W
adm
ar
e
)
3415
m
os
tl
y of
hi
gh
f
r
e
que
nc
y w
or
d pa
ir
s
. T
he
m
e
th
od j
udge
’
s
r
e
c
ti
f
ic
a
ti
on a
nd a
c
c
ur
a
c
y r
a
te
w
hi
c
h i
s
f
ound to be
98 a
nd 96%
r
e
s
pe
c
ti
ve
ly
[
22]
.
N
guye
n
e
t
al
.
[
23]
c
onc
e
nt
r
a
te
s
on
di
f
f
e
r
e
nt
is
s
ue
s
th
a
t
a
r
is
e
du
r
in
g
O
C
R
a
s
a
r
e
s
ul
t
of
in
a
de
qua
te
ly
s
c
a
nne
d
im
a
ge
s
or
li
m
it
a
ti
ons
in
th
e
O
C
R
s
of
twa
r
e
th
a
t
ha
s
be
e
n
de
s
ig
ne
d.
T
he
r
e
s
e
a
r
c
h
pr
opos
e
s
a
hi
ll
c
li
m
bi
ng
a
lg
or
it
hm
-
ba
s
e
d
uns
upe
r
vi
s
e
d
m
ode
l
f
or
O
C
R
e
r
r
or
c
or
r
e
c
ti
on.
C
or
r
e
c
ti
ng
s
ugge
s
ti
ons
a
r
e
r
a
nke
d
a
nd
s
c
or
e
d
us
in
g
a
w
e
ig
ht
e
d
obj
e
c
ti
ve
f
unc
ti
on,
f
or
w
hi
c
h
opt
im
a
l
w
e
ig
ht
c
om
bi
na
ti
ons
a
r
e
de
te
r
m
in
e
d
he
ur
is
ti
c
a
ll
y
[
23]
.
A
f
te
r
e
xt
e
ns
iv
e
li
te
r
a
tu
r
e
s
ur
ve
y,
it
is
f
ound
th
a
t
th
e
r
e
is
a
n
e
e
d
f
or
de
ve
lo
pm
e
nt
of
O
C
R
to
ol
to
id
e
nt
if
y
te
xt
f
r
om
im
a
ge
s
.
T
hi
s
to
ol
is
u
s
e
f
ul
f
or
a
na
ly
s
is
,
a
ut
om
a
ti
on
pr
oc
e
s
s
e
s
a
nd
im
pr
ovi
s
e
d
pr
oduc
ti
vi
ty
f
or
di
f
f
e
r
e
nt
pu
r
pos
e
s
.
A
c
om
pa
r
is
on
ta
bl
e
of
di
f
f
e
r
e
nt
O
C
R
to
ol
s
a
lr
e
a
dy
a
va
il
a
bl
e
b
a
s
e
d
on
th
e
pa
r
a
m
e
te
r
s
s
uc
h a
s
a
c
c
ur
a
c
y, h
a
ndl
in
g c
om
pl
e
x l
a
yout
s
, s
p
e
e
d, e
a
s
e
of
us
e
,
a
nd c
os
t
is
s
how
n i
n T
a
bl
e
1.
F
r
om
th
e
c
om
pa
r
is
on,
it
i
s
e
vi
d
e
nt
th
a
t
T
e
s
s
e
r
a
c
t
O
C
R
of
f
e
r
s
a
n
a
c
c
ur
a
c
y
r
a
nge
of
85
-
99%
,
w
it
h
a
pr
oc
e
s
s
in
g
s
pe
e
d
of
2.5 pa
ge
s
pe
r
s
e
c
ond.
I
t
is
a
f
r
e
e
t
ool
a
nd
c
a
n ha
ndl
e
c
om
pl
e
x
l
a
yout
s
up
t
o 90%
. I
t
is
a
ls
o
r
e
la
ti
ve
ly
e
a
s
y
to
u
s
e
c
om
p
a
r
e
d
to
ot
he
r
O
C
R
t
ool
s
.
T
e
s
s
e
r
a
c
t'
s
a
c
c
ur
a
c
y
i
s
in
f
lu
e
nc
e
d
by
f
a
c
to
r
s
s
uc
h
a
s
im
a
ge
qua
li
ty
, l
a
ngua
ge
,
a
nd nois
e
.
F
or
hi
gh
-
qua
li
ty
pr
in
te
d t
e
xt
, a
c
c
u
r
a
c
y c
a
n
e
xc
e
e
d
95%
, w
he
r
e
a
s
f
or
ha
ndw
r
it
te
n
te
xt
,
lo
w
-
qua
li
ty
or
noi
s
y
im
a
ge
s
,
it
m
a
y
dr
op
to
70
-
85
%
o
r
lo
w
e
r
.
T
hi
s
s
ur
ve
y
h
ig
hl
ig
ht
s
th
e
ne
e
d
f
o
r
a
ddi
ti
ona
l
r
e
s
e
a
r
c
h
in
O
C
R
te
c
hnol
ogi
e
s
to
im
pr
ove
th
e
m
,
e
s
p
e
c
ia
ll
y
to
lo
w
e
r
e
r
r
or
r
a
te
s
,
de
a
l
w
it
h
c
om
pl
e
x
la
yout
s
a
nd e
nha
nc
in
g
s
pe
e
d w
hi
le
m
a
in
ta
in
in
g hi
gh a
c
c
ur
a
c
y.
T
a
bl
e
1. C
om
pa
r
is
on of
di
f
f
e
r
e
nt
O
C
R
t
ool
s
O
C
R
m
e
t
hod
A
c
c
ur
a
c
y
(%)
H
a
ndl
i
ng c
om
pl
e
x l
a
yout
s
(%)
S
pe
e
d (
pa
ge
s
/
s
)
E
a
s
e
of
u
se
C
os
t
T
e
s
s
e
r
a
c
t
O
C
R
[
24]
85
-
99
60
2.5
7/
10
F
r
e
e
E
a
s
yO
C
R
90
-
95
85
1.5
8/
10
F
r
e
e
A
m
a
z
on T
e
xt
r
a
c
t
[
24]
95
-
98
90
1.8
8/
10
$1.50 pe
r
1000 pa
ge
s
A
dobe
A
c
r
oba
t
O
C
R
90
-
95
75
2.5
9/
10
$14.99 pe
r
m
ont
h
O
C
R
.S
pa
c
e
85
-
90
70
2.7
9/
10
F
r
e
e
(
l
i
m
i
t
e
d)
G
oogl
e
D
oc
um
e
nt
A
I
[
24]
95
-
98
90
2.0
9/
10
P
r
i
c
i
ng va
r
i
e
s
(
pa
y
-
as
-
you
-
go)
3.
M
E
T
H
O
D
O
L
O
G
Y
T
he
r
e
s
e
a
r
c
h
p
a
pe
r
f
oc
us
e
s
on
th
e
de
v
e
lo
pm
e
nt
of
a
n
O
C
R
t
ool
us
in
g
c
om
put
e
r
vi
s
io
n
a
nd
N
L
P
.
I
t
is
de
ve
lo
pe
d
in
P
yt
hon
la
ngua
ge
w
it
h
di
f
f
e
r
e
nt
li
br
a
r
ie
s
.
P
yt
hon
li
br
a
r
ie
s
of
c
om
put
e
r
vi
s
io
n
us
e
d
a
r
e
O
pe
nC
V
,
N
um
P
y
a
nd
P
yt
e
s
s
e
r
a
c
t
a
nd
li
br
a
r
ie
s
of
N
L
P
us
e
d
in
c
lu
de
s
pa
C
y
,
P
a
nda
s
,
r
e
gul
a
r
e
xpr
e
s
s
io
n
,
a
nd
s
tr
in
g
.
T
he
e
xt
r
a
c
te
d
te
xt
c
a
n
be
c
ol
le
c
ti
ve
ly
dow
nl
oa
de
d
in
M
ic
r
os
of
t
E
xc
e
l
f
or
e
a
s
e
of
m
a
na
ge
m
e
nt
of
te
xt
e
xt
r
a
c
te
d f
r
om
t
he
i
m
a
ge
s
. T
he
w
or
kf
lo
w
f
or
t
he
de
ve
lo
pm
e
nt
of
a
n
O
C
R
to
ol
is
de
pi
c
te
d i
n t
he
F
ig
ur
e
1.
i)
S
te
p 1:
da
ta
pr
e
pa
r
a
ti
on
‒
C
ol
le
c
t
im
a
ge
s
c
ont
a
in
in
g c
e
r
ti
f
ic
a
te
s
or
t
e
xt
t
o be
pr
oc
e
s
s
e
d.
‒
P
yt
e
s
s
e
r
a
c
t,
a
P
yt
hon wr
a
ppe
r
f
or
G
oogl
e
’
s
T
e
s
s
e
r
a
c
t
O
C
R
e
ng
in
e
, e
xt
r
a
c
ts
t
e
xt
f
r
om
i
m
a
ge
s
.
‒
T
e
xt
e
xt
r
a
c
te
d f
r
om
i
m
a
ge
s
i
s
pr
e
pr
oc
e
s
s
e
d t
o r
e
m
ove
noi
s
e
, f
o
r
m
a
tt
in
g, a
nd i
r
r
e
le
va
nt
da
ta
.
ii)
S
te
p
2:
la
be
li
ng
N
E
R
d
a
ta
‒
N
E
R
da
ta
i
s
l
a
be
ll
e
d m
a
nua
ll
y u
s
in
g t
he
B
I
O
t
a
ggi
ng s
c
he
m
e
.
‒
B
-
B
e
gi
nni
ng:
de
not
e
s
th
e
s
ta
r
t
of
a
n e
nt
it
y.
‒
I
-
I
ns
id
e
:
in
di
c
a
te
s
th
e
c
ont
in
ua
ti
on of
a
n e
nt
it
y.
‒
O
-
O
ut
s
id
e
:
m
a
r
ks
a
r
e
a
s
not
pa
r
t
of
a
ny e
nt
it
y.
iii)
S
te
p 3:
da
ta
pr
e
pr
oc
e
s
s
in
g
‒
T
he
l
a
be
ll
e
d N
E
R
da
t
a
i
s
f
or
m
a
tt
e
d t
o a
li
gn w
it
h
s
pa
C
y
’
s
t
r
a
in
i
ng f
or
m
a
t.
‒
T
he
l
a
be
ll
e
d da
t
a
i
s
c
onve
r
te
d i
nt
o a
f
or
m
a
t
c
om
pa
ti
bl
e
w
it
h
s
p
a
C
y
f
or
N
E
R
m
ode
l
tr
a
in
in
g.
iv
)
S
te
p 4:
N
E
R
m
ode
l
tr
a
in
in
g
‒
D
e
f
in
e
t
he
a
r
c
hi
te
c
tu
r
e
a
nd pa
r
a
m
e
te
r
s
of
t
he
N
E
R
m
ode
l
u
s
in
g
s
pa
C
y
.
‒
T
he
N
E
R
m
ode
l
is
t
r
a
in
e
d on pr
e
pa
r
e
d da
t
a
w
it
h opti
m
iz
a
ti
ons
f
or
pe
r
f
or
m
a
nc
e
.
v)
S
te
p 5:
N
E
R
pr
e
di
c
ti
ons
a
nd da
ta
pi
pe
li
ne
‒
T
he
t
r
a
in
e
d N
E
R
m
ode
l
is
l
oa
de
d t
o m
a
k
e
pr
e
di
c
ti
ons
on ne
w
d
a
ta
.
‒
s
pa
C
y
’
s
di
s
pl
a
C
y
m
odul
e
i
s
ut
il
iz
e
d t
o r
e
nde
r
a
nd
s
e
r
ve
N
E
R
p
r
e
di
c
ti
ons
vi
s
ua
ll
y.
‒
B
ounding boxes
a
r
e
ove
r
la
id
on i
m
a
ge
s
t
o hi
ghl
ig
ht
r
e
c
ogni
z
e
d
e
nt
it
ie
s
.
‒
R
e
c
ogni
z
e
d e
nt
it
ie
s
a
r
e
e
xt
r
a
c
te
d
a
nd pa
r
s
e
d f
r
om
t
he
t
e
xt
f
or
f
ur
th
e
r
pr
oc
e
s
s
in
g or
di
s
pl
a
y.
vi
)
S
te
p 6:
W
e
b A
pp
c
r
e
a
ti
on:
th
e
de
ve
lo
pe
d c
om
pon
e
nt
s
a
r
e
i
nt
e
g
r
a
te
d t
o c
r
e
a
te
a
us
e
r
-
f
r
ie
ndl
y w
e
b
a
ppl
ic
a
ti
on a
ll
ow
in
g us
e
r
s
t
o uploa
d c
e
r
ti
f
ic
a
te
i
m
a
ge
s
, e
xt
r
a
c
t
t
e
xt
, i
de
nt
if
y e
nt
it
ie
s
, a
nd vis
ua
li
z
e
t
he
r
e
s
ul
ts
e
f
f
e
c
ti
ve
ly
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 4,
A
ugus
t
2025
:
3412
-
3420
3416
vi
i)
S
te
p
7:
T
he
te
xt
obt
a
in
e
d
f
r
om
th
e
e
xt
r
a
c
ti
on
pr
oc
e
s
s
is
s
ub
s
e
qu
e
nt
ly
s
to
r
e
d
w
it
hi
n
an
E
xc
e
l
f
il
e
.
D
if
f
e
r
e
nt
li
br
a
r
ie
s
,
th
e
ir
f
unc
ti
ons
a
nd
th
e
ir
a
lg
or
it
hm
s
us
e
d
in
th
e
de
ve
lo
pm
e
nt
of
a
n
O
C
R
to
ol
a
r
e
:
C
a
nny
e
dge
d
e
te
c
ti
on
a
lg
or
it
hm
a
nd
D
ougl
a
s
-
P
e
uc
ke
r
a
lg
or
it
hm
.
C
a
nny
e
dge
d
e
te
c
ti
on
a
lg
or
it
hm
c
om
put
e
s
th
e
gr
a
di
e
nt
m
a
gni
tu
de
a
nd
di
r
e
c
ti
on
f
or
e
a
c
h
pi
xe
l
us
in
g
te
c
hni
que
s
s
uc
h
a
s
S
obe
l
ope
r
a
to
r
s
.
I
t
th
e
n
s
uppr
e
s
s
e
s
non
-
m
a
xi
m
um
gr
a
di
e
nt
va
lu
e
s
to
th
in
out
de
te
c
te
d
e
dge
s
,
r
e
ta
in
in
g
onl
y
th
e
lo
c
a
l
m
a
xi
m
a
a
lo
ng
th
e
e
dge
s
.
I
t
f
ol
lo
w
s
e
dge
s
by
li
nki
ng
a
dj
a
c
e
nt
pi
xe
ls
w
it
h
gr
a
di
e
nt
m
a
gni
tu
de
s
a
bove
a
hi
gh
th
r
e
s
hol
d
a
nd
pot
e
nt
ia
ll
y
w
e
a
k
e
dge
s
a
bove
a
lo
w
th
r
e
s
hol
d
a
nd
de
te
r
m
in
e
w
hi
c
h
w
e
a
k
e
dg
e
s
to
r
e
ta
in
ba
s
e
d
on
th
e
ir
c
onn
e
c
ti
vi
ty
to
s
tr
ong
e
dge
s
. D
ougl
a
s
-
P
e
uc
k
e
r
a
lg
or
it
hm
id
e
nt
if
ie
s
a
pol
yl
in
e
de
f
in
e
d
by a
s
e
que
nc
e
of
poi
nt
s
i
n a
pl
a
ne
. T
h
e
out
put
is
th
a
t
th
e
a
lg
or
it
hm
ge
ne
r
a
te
s
a
s
im
pl
if
ie
d
pol
yl
in
e
by
r
e
ta
in
in
g
c
r
it
ic
a
l
poi
nt
s
th
a
t
de
f
in
e
th
e
s
h
a
pe
a
c
c
ur
a
te
ly
.
F
ig
ur
e
1. W
or
ki
ng of
O
C
R
t
ool
4.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
S
T
he
O
C
R
to
ol
i
s
s
p
e
c
if
ic
a
ll
y
c
r
a
f
te
d
to
e
xt
r
a
c
t
te
xt
f
r
om
i
m
a
ge
s
a
nd
s
to
r
e
th
e
obt
a
in
e
d
te
xt
in
M
ic
r
os
of
t
E
xc
e
l.
I
t
is
tr
a
in
e
d
on
th
e
da
ta
s
e
t
of
ove
r
8
,
000
im
a
ge
s
w
hi
c
h
in
c
lu
de
s
6
,
500
im
a
ge
s
u
s
e
d
f
or
tr
a
in
in
g
,
a
nd 1
,
500 im
a
ge
s
us
e
d f
or
t
e
s
ti
ng. T
he
s
te
p
-
by
-
s
te
p pr
oc
e
s
s
t
o
us
e
O
C
R
to
ol
i
s
gi
ve
n
a
s
f
ol
lo
w
:
i)
S
te
p
1:
upl
oa
d
th
e
im
a
ge
f
or
te
xt
e
xt
r
a
c
ti
on
pr
oc
e
s
s
by c
li
c
ki
ng
“
U
pl
oa
d
I
m
a
ge
”
but
to
n
a
nd
c
li
c
k “
W
r
a
p
C
e
r
ti
f
ic
a
te
a
nd
E
xt
r
a
c
t
T
e
xt
”
.
A
f
te
r
th
a
t,
s
e
le
c
t
th
e
bounda
r
ie
s
o
f
th
e
im
a
ge
in
di
c
a
ti
ng
th
e
pa
r
t
f
r
o
m
w
he
r
e
th
e
t
e
xt
i
s
t
o be
e
xt
r
a
c
te
d.
ii)
S
te
p 2:
e
xt
r
a
c
te
d
te
xt
f
r
om
l
oa
de
d i
m
a
ge
i
s
s
how
n i
n t
a
bl
e
f
or
m
a
t
a
s
s
how
n i
n F
ig
ur
e
2.
iii)
S
te
p
3:
c
li
c
k
“
D
ow
nl
oa
d
a
s
E
xc
e
l”
to
dow
nl
oa
d
e
xt
r
a
c
te
d
te
x
t
in
E
xc
e
l
f
or
m
a
t
a
nd
c
li
c
k
th
e
“
B
a
c
k
to
H
om
e
”
but
to
n t
o go ba
c
k t
o m
a
in
hom
e
pa
ge
a
s
s
how
n i
n F
ig
ur
e
3.
F
ig
ur
e
2. T
e
xt
e
xt
r
a
c
ti
on f
r
om
i
m
a
ge
i
n t
a
bl
e
f
or
m
a
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
T
r
ans
fo
r
m
in
g i
m
age
s
i
nt
o w
or
d
s
:
opt
ic
al
c
har
ac
te
r
r
e
c
ogni
ti
on s
ol
ut
io
ns
f
or
…
(
J
y
ot
i
W
adm
ar
e
)
3417
F
ig
ur
e
3. R
e
por
t
ge
ne
r
a
ti
on i
n E
xc
e
l
f
or
m
a
t
4.1. Re
s
u
lt
val
id
at
io
n
of
O
C
R
t
ool
T
he
a
c
c
ur
a
c
y
of
th
e
O
C
R
to
ol
i
s
e
v
a
lu
a
t
e
d
u
s
in
g
two
p
a
r
a
m
e
te
r
s
:
C
E
R
a
nd
w
or
d
e
r
r
or
r
a
te
(
W
E
R
)
[
25]
.
T
he
C
E
R
i
s
u
s
e
d
to
c
a
lc
ul
a
t
e
th
e
to
ta
l
c
h
a
r
a
c
t
e
r
c
o
unt
th
a
t
i
s
m
is
m
a
t
c
h
e
d
or
d
e
te
c
te
d
i
nc
or
r
e
c
t
d
ur
in
g
te
xt
e
xt
r
a
c
ti
on
to
t
h
e
t
ot
a
l
numb
e
r
of
c
h
a
r
a
c
t
e
r
s
i
n
th
e
or
ig
in
a
l
t
e
xt
.
I
t
i
s
us
e
d t
o m
e
a
s
ur
e
t
he
c
o
unt
of
c
h
a
r
a
c
te
r
s
t
h
a
t
a
r
e
s
ub
s
ti
tu
t
e
d, d
e
le
te
d
or
i
nc
or
r
e
c
t
c
h
a
r
a
c
t
e
r
s
i
n
s
e
r
te
d dur
in
g t
h
e
t
e
xt
e
xt
r
a
c
ti
on pr
o
c
e
s
s
. C
E
R
i
s
gi
v
e
n by
(
1)
.
=
(
ℎ
)
+
(
ℎ
)
+
(
ℎ
)
(
ℎ
)
(
1)
W
he
r
e
S
(
c
ha
r
)
s
ta
nds
f
or
to
ta
l
c
ha
r
a
c
te
r
c
ount
th
a
t
a
r
e
s
ubs
ti
t
ut
e
d
f
r
om
th
e
or
ig
in
a
l
te
xt
;
I
(
c
ha
r
)
de
not
e
th
e
num
be
r
of
in
c
or
r
e
c
t
c
ha
r
a
c
te
r
s
th
a
t
a
r
e
in
s
e
r
te
d
in
th
e
e
xt
r
a
c
te
d
t
e
xt
;
D
(
c
ha
r
)
s
ig
ni
f
ie
s
th
e
num
be
r
of
c
ha
r
a
c
t
e
r
s
not
r
e
c
ogni
z
e
d
or
m
is
s
in
g
in
th
e
e
xt
r
a
c
te
d
te
xt
;
a
nd
N
(
c
ha
r
)
in
di
c
a
te
s
th
e
to
t
a
l
c
ha
r
a
c
te
r
c
ount
pr
e
s
e
nt
in
th
e
or
ig
in
a
l
te
xt
.
T
he
ty
pi
c
a
l
C
E
R
s
houl
d be
i
n t
he
r
a
ng
e
of
2
-
10%
. T
he
C
E
R
of
O
C
R
t
ool
de
v
e
lo
pe
d i
s
2.4%
.
T
he
W
E
R
is
us
e
d
to
c
a
l
c
ul
a
te
t
he
t
ot
a
l
num
be
r
of
w
or
d
s
th
a
t
a
r
e
m
i
s
m
a
t
c
he
d
or
d
e
te
c
te
d
in
c
or
r
e
c
t
dur
in
g t
e
xt
e
xt
r
a
c
ti
on t
o t
h
e
t
ot
a
l
nu
m
be
r
of
w
or
d
s
i
n t
he
or
ig
in
a
l
te
xt
. I
t
is
u
s
e
d t
o m
e
a
s
ur
e
t
he
c
o
unt
of
w
or
d
s
th
a
t
a
r
e
s
ub
s
ti
tu
t
e
d,
de
l
e
te
d
or
i
n
c
or
r
e
c
t
w
or
ds
in
s
e
r
te
d
dur
in
g
t
h
e
t
e
x
t
e
x
tr
a
c
ti
on pr
o
c
e
s
s
.
W
E
R
i
s
gi
ve
n
by
(
2)
.
=
(
)
+
(
)
+
(
)
(
)
(
2)
W
he
r
e
S
(
w
or
d)
s
ta
nds
f
or
th
e
num
be
r
of
w
or
ds
th
a
t
a
r
e
s
ubs
ti
tu
te
d
f
r
om
th
e
or
ig
in
a
l
te
xt
;
I
(
w
or
d
)
de
not
e
th
e
num
be
r
of
in
c
or
r
e
c
t
w
or
ds
th
a
t
a
r
e
in
s
e
r
te
d
in
th
e
e
xt
r
a
c
te
d
t
e
xt
;
D
(
w
or
d)
s
ig
ni
f
ie
s
th
e
num
be
r
of
w
or
ds
not
r
e
c
ogni
z
e
d
or
m
is
s
in
g
in
th
e
e
xt
r
a
c
te
d
te
xt
;
a
nd
N
(
w
or
d)
in
d
i
c
a
te
s
to
ta
l
w
or
ds
c
ount
pr
e
s
e
nt
in
th
e
or
ig
in
a
l
te
xt
. T
he
ta
r
ge
t
W
E
R
s
houl
d be
l
e
s
s
t
ha
n 5%
.
T
he
W
E
R
of
O
C
R
t
ool
de
ve
lo
pe
d i
s
2.8%
.
T
he
a
c
c
ur
a
c
y of
t
he
ove
r
a
ll
O
C
R
to
ol
de
ve
lo
pe
d i
s
98.8%
.
4.2 Com
p
ar
is
on
w
it
h
e
xi
s
t
in
g s
ys
t
e
m
T
a
bl
e
2
c
om
pa
r
e
s
va
r
io
us
O
C
R
a
ppr
oa
c
he
s
a
nd
th
e
ir
a
c
c
ur
a
c
ie
s
on
di
f
f
e
r
e
nt
ty
pe
s
of
te
xt
.
I
t
c
om
bi
ne
s
s
ta
nda
r
d
O
C
R
a
lg
or
it
hm
s
,
c
u
s
to
m
-
bui
lt
m
ode
ls
,
a
nd
tr
a
ns
f
or
m
e
d
m
a
c
hi
ne
le
a
r
ni
ng
te
c
hni
que
s
.
T
he
pe
r
f
or
m
a
nc
e
m
e
a
s
ur
e
m
e
nt
s
in
c
lu
de
ty
pe
w
r
it
te
n
a
nd
ha
ndw
r
it
t
e
n
te
xt
,
a
s
w
e
ll
a
s
s
pe
c
if
ic
da
ta
s
e
ts
.
T
a
bl
e
2
de
m
ons
tr
a
te
s
t
h
e
da
ta
s
e
t
u
s
e
d
a
nd a
c
c
ur
a
c
y of
e
a
c
h
O
C
R
t
e
c
hni
que
. T
r
a
di
ti
ona
l
a
ppr
oa
c
he
s
, s
u
c
h a
s
T
e
s
s
e
r
a
c
t
O
C
R
,
a
c
hi
e
ve
good
a
c
c
ur
a
c
y
f
or
ty
pe
w
r
it
te
n
te
xt
,
a
lt
hough
s
ophi
s
ti
c
a
te
d
m
od
e
ls
w
it
h
U
-
N
e
t
a
nd
C
R
N
N
s
tr
u
c
tu
r
e
s
pe
r
f
or
m
c
om
pe
ti
ti
ve
ly
.
T
he
s
e
in
s
ig
ht
s
he
lp
c
hoo
s
e
th
e
be
s
t
O
C
R
t
e
c
hni
que
ba
s
e
d
on
pa
r
ti
c
ul
a
r
a
ppl
ic
a
ti
on ne
e
ds
a
nd t
e
xt
c
ha
r
a
c
te
r
is
ti
c
s
.
T
a
bl
e
2. S
um
m
a
r
y of
d
if
f
e
r
e
nt
O
C
R
te
c
hni
que
s
a
nd t
he
ir
a
c
c
ur
a
c
ie
s
T
e
c
hni
que
/
m
ode
l
us
e
d
D
a
t
a
s
e
t
A
c
c
ur
a
c
y
S
l
a
nt
c
or
r
e
c
t
i
on
l
a
ye
r
a
nd
c
ha
r
a
c
t
e
r
s
e
gm
e
nt
a
t
i
on a
nd r
e
c
ogni
t
i
on
[
3]
I
C
D
A
R
2013
-
1
,
081 i
m
a
ge
s
S
e
l
f
m
a
de
-
8
,
000 i
m
a
ge
s
96.42%
-
I
C
D
A
R
2013 da
t
a
s
e
t
.
96.52%
-
s
e
l
f
-
m
a
de
s
c
r
e
e
n r
e
nde
r
e
d
da
t
a
s
e
t
U
-
N
e
t
i
m
a
ge
s
e
gm
e
nt
a
t
i
on
[
8]
55
,
000 i
m
a
ge
s
98%
R
oot
S
I
F
T
w
i
t
h
A
B
C
opt
i
m
i
z
e
d
ne
ur
a
l
ne
t
w
or
k a
l
gor
i
t
hm
[
12]
500 t
r
a
i
ni
ng i
m
a
ge
s
97.31%
S
t
a
t
i
s
t
i
c
a
l
c
l
a
s
s
i
f
i
c
a
t
i
on a
ppr
oa
c
h
e
s
[
18]
I
C
D
A
R
2013
-
1
,
081 i
m
a
ge
s
T
ype
w
r
i
t
t
e
n t
e
xt
:
97%
H
a
ndw
r
i
t
t
e
n t
e
xt
:
80 t
o 90%
C
N
N
w
i
t
h R
N
N
[
24]
10
,
419 i
m
a
ge
s
96.21%
T
e
s
s
e
r
a
c
t
O
C
R
-
our
m
e
t
hod
8
,
000 i
m
a
ge
s
T
ype
w
r
i
t
t
e
n t
e
xt
:
98.8%
H
a
ndw
r
i
t
t
e
n t
e
xt
:
90.6%
5.
C
O
N
C
L
U
S
I
O
N
A
N
D
F
U
T
U
R
E
S
C
O
P
E
T
he
O
C
R
to
ol
pl
a
y
s
a
c
r
uc
ia
l
r
ol
e
in
e
xt
r
a
c
ti
ng
te
xt
f
r
om
im
a
ge
s
,
im
pr
ovi
ng
e
f
f
ic
ie
nc
y
in
va
r
io
us
in
du
s
tr
i
e
s
by
c
onv
e
r
ti
n
g
s
c
a
n
s
, i
m
a
g
e
s
,
a
nd
ha
nd
w
r
it
in
g
in
to
e
di
t
a
bl
e
f
or
m
a
t
s
.
R
e
c
e
nt
a
dv
a
nc
e
m
e
nt
s
in
m
a
c
h
in
e
le
a
r
ni
ng
a
nd
c
om
put
e
r
vi
s
io
n
of
f
e
r
e
nha
nc
e
d
a
c
c
ur
a
c
y,
p
a
r
ti
c
ul
a
r
ly
in
c
ha
ll
e
ngi
ng
s
c
e
n
a
r
io
s
li
ke
noi
s
y
im
a
ge
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 4,
A
ugus
t
2025
:
3412
-
3420
3418
or
c
om
pl
e
x
f
ont
s
.
B
r
oa
de
ni
ng
la
ngua
g
e
s
uppor
t
w
il
l
f
ur
th
e
r
e
nh
a
nc
e
th
e
c
a
pa
bi
li
ti
e
s
of
O
C
R
de
vi
c
e
s
to
ha
ndl
e
a
va
r
ie
ty
of
s
c
r
ip
ts
.
B
y
in
te
gr
a
ti
ng
O
C
R
w
it
h
e
m
e
r
gi
ng
te
c
hnol
ogi
e
s
li
ke
I
oT
a
nd
a
ugm
e
nt
e
d
r
e
a
li
ty
(
AR
)
,
a
nd
tr
a
in
in
g
it
to
de
te
c
t
ha
ndw
r
it
te
n
a
nd
c
ur
s
iv
e
te
xt
,
it
s
ve
r
s
a
ti
li
ty
is
in
c
r
e
a
s
e
d.
A
c
c
e
s
s
ib
il
it
y
f
e
a
tu
r
e
s
s
uc
h
a
s
voi
c
e
c
ont
r
ol
a
nd
s
c
r
e
e
n
r
e
a
d
e
r
c
om
pa
ti
bi
li
ty
c
a
te
r
to
a
w
id
e
r
a
ng
e
of
us
e
r
s
,
in
c
lu
di
ng
th
os
e
w
it
h
di
s
a
bi
li
ti
e
s
.
I
m
pl
e
m
e
nt
in
g
r
obus
t
s
e
c
u
r
it
y m
e
a
s
ur
e
s
,
in
c
lu
di
ng
e
nc
r
ypt
io
n
a
nd
c
om
pl
ia
nc
e
w
it
h
r
e
gul
a
ti
ons
,
is
e
s
s
e
nt
ia
l
f
or
pr
ot
e
c
ti
ng
s
e
ns
it
iv
e
da
ta
.
I
n
c
on
c
lu
s
io
n,
th
e
f
ut
ur
e
of
O
C
R
to
o
ls
lo
oks
pr
om
is
in
g,
dr
iv
e
n
by
in
nova
ti
on
a
nd
us
e
r
r
e
qui
r
e
m
e
nt
s
.
E
m
pha
s
iz
in
g
a
c
c
ur
a
c
y,
la
ngu
a
ge
di
ve
r
s
it
y,
in
te
gr
a
ti
on,
c
us
to
m
iz
a
ti
on,
a
c
c
e
s
s
ib
il
it
y,
a
nd
s
e
c
ur
it
y
w
il
l
e
ns
ur
e
th
a
t
O
C
R
to
ol
s
c
ont
in
ue
to
e
vol
ve
a
s
in
di
s
pe
ns
a
bl
e
s
ol
ut
io
n
s
f
or
te
xt
pr
oc
e
s
s
in
g
in
th
e
di
gi
ta
l
la
nds
c
a
pe
.
A
C
K
N
O
WL
E
D
G
M
E
N
T
S
T
he
a
ut
hor
s
f
ul
ly
a
c
knowle
dge
t
he
i
ns
ti
tu
te
, K
.
J
. S
om
a
iy
a
I
ns
t
it
ut
e
of
T
e
c
hnol
ogy f
o
r
i
ts
s
uppor
t
on
th
is
r
e
s
e
a
r
c
h.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
A
ut
hor
s
s
ta
te
no f
undi
ng i
nvol
ve
d.
A
U
T
H
O
R
C
O
N
T
R
I
B
U
T
I
O
N
S
S
T
A
T
E
M
E
N
T
T
hi
s
jo
ur
na
l
us
e
s
th
e
C
ont
r
ib
ut
or
R
ol
e
s
T
a
xonomy
(
C
R
e
di
T
)
to
r
e
c
ogni
z
e
in
di
vi
dua
l
a
ut
hor
c
ont
r
ib
ut
io
ns
, r
e
duc
e
a
ut
hor
s
hi
p di
s
put
e
s
,
a
nd f
a
c
il
it
a
te
c
ol
la
bo
r
a
ti
on.
N
am
e
o
f
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
J
yot
i
W
a
dm
a
r
e
✓
✓
✓
✓
✓
✓
✓
✓
✓
S
uni
ta
R
a
vi
ndr
a
P
a
ti
l
✓
✓
✓
✓
✓
✓
D
a
ks
hi
ta
K
ol
te
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
K
a
pi
l
B
ha
ti
a
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
P
a
la
k D
e
s
a
i
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
G
a
ne
s
h W
a
dm
a
r
e
✓
✓
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
l
i
z
a
t
i
on
M
:
M
e
t
hodol
ogy
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
da
t
i
on
Fo
:
Fo
r
m
a
l
a
na
l
ys
i
s
I
:
I
nve
s
t
i
ga
t
i
on
R
:
R
e
s
our
c
e
s
D
:
D
a
t
a
C
ur
a
t
i
on
O
:
W
r
i
t
i
ng
-
O
r
i
gi
na
l
D
r
a
f
t
E
:
W
r
i
t
i
ng
-
R
e
vi
e
w
&
E
di
t
i
ng
Vi
:
Vi
s
ua
l
i
z
a
t
i
on
Su
:
Su
pe
r
vi
s
i
on
P
:
P
r
oj
e
c
t
a
dm
i
ni
s
t
r
a
t
i
on
Fu
:
Fu
ndi
ng
a
c
qui
s
i
t
i
on
C
O
N
F
L
I
C
T
O
F
I
N
T
E
R
E
S
T
S
T
A
T
E
M
E
N
T
A
ut
hor
s
s
ta
te
no c
onf
li
c
t
of
i
nt
e
r
e
s
t.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
T
he
da
ta
th
a
t
s
uppor
t
th
e
f
in
di
ngs
of
th
is
s
tu
dy
a
r
e
a
va
il
a
bl
e
f
r
o
m
th
e
c
or
r
e
s
ponding
a
ut
hor
,
[
J
W
]
,
upo
n
r
e
a
s
ona
bl
e
r
e
que
s
t.
R
E
F
E
R
E
N
C
E
S
[
1]
P
.
J
a
i
n,
D
.
K
. T
a
ne
j
a
,
a
nd
D
.
H
. T
a
ne
j
a
,
“
W
hi
c
h
O
C
R
t
ool
s
e
t
i
s
good
a
nd
w
hy
?
a
c
om
pa
r
a
t
i
ve
s
t
udy,”
K
uw
ai
t
J
our
nal
of
Sc
i
e
nc
e
,
vol
. 48, no. 2, A
pr
. 2021, doi
:
10.48129/
kj
s
.v48i
2.9589.
[
2]
W
.
S
un,
L
.
L
i
u,
W
.
Z
ha
ng,
a
nd
J
.
C
.
C
om
f
or
t
,
“
I
nt
e
l
l
i
ge
nt
O
C
R
pr
oc
e
s
s
i
ng,”
J
our
nal
of
t
he
A
m
e
r
i
c
an
So
c
i
e
t
y
f
o
r
I
nf
or
m
at
i
on
Sc
i
e
nc
e
, vol
. 43, no. 6, pp. 422
–
431, 1992, doi
:
10.1002/
(
S
I
C
I
)
1097
-
4571
(
1992
07)
43:
63.0.C
O
;
2
-
Z.
[
3]
T
. T
.
H
. N
guye
n, A
.
J
a
t
ow
t
, M
. C
ous
t
a
t
y, a
nd A
. D
ouc
e
t
, “
S
u
r
ve
y of
pos
t
-
O
C
R
pr
oc
e
s
s
i
ng a
ppr
oa
c
he
s
,
”
A
C
M
C
om
put
i
ng Sur
v
e
y
s
,
vol
. 54, no. 6, pp. 1
–
37, J
ul
. 2022, doi
:
10.1145/
3453476.
[
4]
T
.
H
e
ggha
m
m
e
r
,
“
O
C
R
w
i
t
h
T
e
s
s
e
r
a
c
t
,
A
m
a
z
on
T
e
xt
r
a
c
t
,
a
nd
G
oogl
e
doc
um
e
nt
A
I
:
a
be
nc
hm
a
r
ki
ng
e
xpe
r
i
m
e
nt
,”
J
our
nal
of
C
om
put
at
i
onal
Soc
i
al
Sc
i
e
nc
e
, vol
. 5, no. 1, pp. 861
–
882, M
a
y 2022, doi
:
10.10
07/
s
42001
-
021
-
00149
-
1.
[
5]
O
.
O
.
O
l
a
da
yo,
“
O
pt
i
c
a
l
c
ha
r
a
c
t
e
r
r
e
c
ogni
t
i
on
of
o
f
f
-
l
i
ne
t
ype
d
a
nd
ha
ndw
r
i
t
t
e
n
E
ngl
i
s
h
t
e
xt
us
i
ng
m
or
phol
ogi
c
a
l
a
nd
t
e
m
pl
a
t
e
m
a
t
c
hi
ng
t
e
c
hni
que
s
,”
I
A
E
S
I
nt
e
r
nat
i
onal
J
our
nal
of
A
r
t
i
f
i
c
i
al
I
n
t
e
l
l
i
ge
nc
e
,
vol
.
3,
no.
3,
p
p
.
121
-
128
,
S
e
p.
2014,
doi
:
10.11591/
i
j
a
i
.v3.i
3.pp121
-
128.
[
6]
C
.
A
dj
e
t
e
y
a
nd
K
.
S
.
A
.
-
M
a
n
u,
“
C
ont
e
n
t
-
ba
s
e
d
i
m
a
ge
r
e
t
r
i
e
va
l
us
i
n
g
T
e
s
s
e
r
a
c
t
O
C
R
e
ngi
ne
a
nd
l
e
ve
ns
h
t
e
i
n
a
l
g
or
i
t
hm
,”
I
nt
e
r
n
at
i
o
nal
J
our
nal
o
f
A
dv
anc
e
d C
o
m
put
e
r
Sc
i
e
nc
e
a
nd
A
ppl
i
c
at
i
ons
, vo
l
. 12
, no.
7, 2
021,
d
oi
:
10
.1456
9/
I
J
A
C
S
A
.202
1.012
0776.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
T
r
ans
fo
r
m
in
g i
m
age
s
i
nt
o w
or
d
s
:
opt
ic
al
c
har
ac
te
r
r
e
c
ogni
ti
on s
ol
ut
io
ns
f
or
…
(
J
y
ot
i
W
adm
ar
e
)
3419
[
7]
B
.
Z
hu,
H
.
Z
ha
ng,
W
.
C
he
n,
F
.
X
i
a
,
a
nd
R
.
M
a
c
i
e
j
e
w
s
ki
,
“
S
hot
V
i
s
:
s
m
a
r
t
ph
one
-
ba
s
e
d
vi
s
u
a
l
i
z
a
t
i
on
of
O
C
R
i
nf
or
m
a
t
i
on
f
r
om
i
m
a
ge
s
,”
A
C
M
T
r
an
s
ac
t
i
ons
on
M
ul
t
i
m
e
di
a
C
om
put
i
ng,
C
o
m
m
uni
c
at
i
ons
,
and
A
ppl
i
c
at
i
ons
,
vol
.
12,
no.
1
s
,
pp.
1
–
17,
O
c
t
.
2015
,
doi
:
10.1145/
2808210.
[
8]
G
.
S
uddul
a
nd
J
.
F
.
L
.
S
e
gui
n,
“
A
c
u
s
t
om
-
bui
l
t
de
e
p
l
e
a
r
ni
ng a
ppr
oa
c
h
f
or
t
e
xt
e
xt
r
a
c
t
i
on
f
r
om
i
de
nt
i
t
y
c
a
r
d
i
m
a
ge
s
,”
I
nt
e
r
nat
i
ona
l
J
our
nal
of
I
nf
or
m
at
i
c
s
and
C
om
m
uni
c
at
i
on
T
e
c
hnol
ogy
(
I
J
-
I
C
T
)
,
vo
l
.
13,
no.
1,
p
p.
34
-
41
,
A
pr
.
2024,
doi
:
10.11591/
i
j
i
c
t
.v13i
1.pp34
-
41.
[
9]
K
.
S
a
t
i
r
a
pi
w
ong
a
nd
T
.
S
i
r
i
bo
r
vor
nr
a
t
a
na
kul
,
“
I
n
f
or
m
a
t
i
on
e
xt
r
a
c
t
i
on
f
or
di
f
f
e
r
e
nt
l
a
yout
s
of
i
nvoi
c
e
i
m
a
ge
s
,”
T
he
I
m
agi
ng
Sc
i
e
nc
e
J
our
nal
, vol
. 69, no. 5
–
8, pp. 417
–
429, N
ov. 2021, doi
:
10.1080/
13682199.2022.2157367.
[
10]
M
.
C
.
L
e
e
,
“
I
m
pr
ovi
ng
a
c
c
e
s
s
i
bi
l
i
t
y
i
n
i
nt
e
r
l
i
br
a
r
y
L
oa
n
us
i
ng
O
C
R
,”
J
ou
r
nal
of
I
nt
e
r
l
i
br
ar
y
L
oan,
D
oc
um
e
nt
D
e
l
i
v
e
r
y
&
E
l
e
c
t
r
oni
c
R
e
s
e
r
v
e
, vol
. 29, no. 1
–
2, pp. 75
–
87,
M
a
r
. 2020, doi
:
10.1080/
1072303X
.2020.1859426.
[
11]
P
.
M
a
ni
va
nna
n
e
t
al
.
,
“
D
oc
t
or
unp
r
e
di
c
t
e
d
pr
e
s
c
r
i
pt
i
on
ha
ndw
r
i
t
i
ng
pr
e
di
c
t
i
on
us
i
ng
t
r
i
boe
l
e
c
t
r
i
c
s
m
a
r
t
r
e
c
ogni
t
i
on,”
P
r
oduc
t
i
on
P
l
anni
ng &
C
ont
r
ol
, pp. 1
–
17, A
pr
. 2023, doi
:
10.1080/
09537287.2023.2202173.
[
12]
S
.
B
.
P
oodi
kka
l
a
m
a
nd
P
.
L
oga
na
t
ha
n,
“
O
pt
i
c
a
l
c
ha
r
a
c
t
e
r
r
e
c
ogni
t
i
on
ba
s
e
d
on
l
oc
a
l
i
nva
r
i
a
nt
f
e
a
t
ur
e
s
,”
T
he
I
m
agi
ng
Sc
i
e
nc
e
J
our
nal
, vol
. 68, no. 4, pp. 214
–
224, M
a
y 2020, doi
:
10.1080/
13682199.2020.1
827814.
[
13]
M
.
M
ohd,
F
.
Q
a
m
a
r
,
I
.
A
l
-
S
he
i
kh,
a
nd
R
.
S
a
l
a
h,
“
Q
ur
a
ni
c
opt
i
c
a
l
t
e
xt
r
e
c
og
ni
t
i
on
us
i
ng
de
e
p
l
e
a
r
ni
ng
m
ode
l
s
,”
I
E
E
E
A
c
c
e
s
s
,
vol
. 9, pp.
38318
–
38330, 2021, doi
:
10.1109/
A
C
C
E
S
S
.2021.3064019.
[
14]
H
. H
a
s
s
a
n, A
.
E
l
-
M
a
hdy, a
nd
M
. E
. H
u
s
s
e
i
n,
“
A
r
a
bi
c
s
c
e
ne
t
e
xt
r
e
c
ogni
t
i
on i
n
t
he
de
e
p l
e
a
r
ni
ng e
r
a
:
a
na
l
y
s
i
s
on
a
nove
l
da
t
a
s
e
t
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 9, pp. 107046
–
107058, 2021, doi
:
10.1109/
A
C
C
E
S
S
.2021.3100717.
[
15]
R
.
M
a
l
hot
r
a
a
nd
M
.
T
.
A
ddi
s
,
“
E
nd
-
to
-
e
nd
hi
s
t
or
i
c
a
l
ha
ndw
r
i
t
t
e
n
E
t
hi
opi
c
t
e
xt
r
e
c
ogni
t
i
on
us
i
ng
de
e
p
l
e
a
r
ni
ng,”
I
E
E
E
A
c
c
e
s
s
,
vol
. 11, pp. 99535
–
99545, 2023, doi
:
10.1109/
A
C
C
E
S
S
.2023.3314334.
[
16]
B
.
W
a
ng,
Y
.
W
.
M
a
,
a
nd
H
.
T
.
H
u,
“
H
ybr
i
d
m
ode
l
f
or
C
hi
ne
s
e
c
h
a
r
a
c
t
e
r
r
e
c
ogni
t
i
on
ba
s
e
d
on
T
e
s
s
e
r
a
c
t
-
O
C
R
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
I
nt
e
r
ne
t
P
r
ot
oc
ol
T
e
c
hnol
ogy
, vol
. 13, no. 2, 2020, doi
:
10.1504/
I
J
I
P
T
.2020.106316.
[
17]
K
.
C
.
S
ha
hi
r
a
a
nd
A
.
L
i
j
i
ya
,
“
T
ow
a
r
ds
a
s
s
i
s
t
i
ng
t
he
vi
s
ua
l
l
y
i
m
pa
i
r
e
d:
a
r
e
vi
e
w
on
t
e
c
hni
que
s
f
or
de
c
odi
ng
t
he
vi
s
ua
l
da
t
a
f
r
om
c
ha
r
t
i
m
a
ge
s
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 9, pp. 52926
–
52943, 2021, doi
:
10.1109/
A
C
C
E
S
S
.2021.3069205.
[
18]
G
.
P
ol
a
nc
i
c
,
S
.
J
a
ge
c
i
c
,
a
nd
K
.
K
ous
,
“
A
n
e
m
pi
r
i
c
a
l
i
nve
s
t
i
ga
t
i
on
of
t
he
e
f
f
e
c
t
i
v
e
ne
s
s
of
opt
i
c
a
l
r
e
c
ogni
t
i
on
of
ha
nd
-
dr
a
w
n
bus
i
ne
s
s
pr
oc
e
s
s
e
l
e
m
e
nt
s
by
a
ppl
yi
ng
m
a
c
hi
ne
l
e
a
r
ni
ng,”
I
E
E
E
A
c
c
e
s
s
,
vol
.
8,
pp.
206118
–
206131,
2020,
doi
:
10.1109/
A
C
C
E
S
S
.2020.3034603.
[
19]
A
.
U
e
da
,
W
.
Y
a
ng,
a
nd
K
.
S
ugi
ur
a
,
“
S
w
i
t
c
hi
ng
t
e
xt
-
ba
s
e
d
i
m
a
g
e
e
nc
od
e
r
s
f
or
c
a
pt
i
oni
ng
i
m
a
ge
s
w
i
t
h
t
e
xt
,”
I
E
E
E
A
c
c
e
s
s
,
vol
. 11, pp. 55706
–
55715, 2023, doi
:
10.1109/
A
C
C
E
S
S
.2023.3282444.
[
20]
L
. W
u, Y
.
X
u, J
. H
ou, C
.
L
. P
.
C
he
n, a
nd
C
.
-
L
.
L
i
u, “
A
t
w
o
-
l
e
ve
l
r
e
c
t
i
f
i
c
a
t
i
on
a
t
t
e
nt
i
on ne
t
w
or
k f
or
s
c
e
ne
t
e
xt
r
e
c
ogni
t
i
on,”
I
E
E
E
T
r
ans
ac
t
i
ons
on M
ul
t
i
m
e
di
a
, vol
. 25, pp. 2404
–
2414, 2023, doi
:
10.1109/
T
M
M
.2022.3146779.
[
21]
Y
. Z
ha
ng, S
. N
i
e
, S
. L
i
a
ng, a
nd W
.
L
i
u, “
R
obus
t
t
e
xt
i
m
a
ge
r
e
c
ogni
t
i
on vi
a
a
dv
e
r
s
a
r
i
a
l
s
e
que
nc
e
-
to
-
s
e
que
nc
e
dom
a
i
n a
d
a
pt
a
t
i
on,”
I
E
E
E
T
r
ans
ac
t
i
ons
on I
m
age
P
r
oc
e
s
s
i
ng
, vol
. 30, pp. 3922
–
3933, 2021, doi
:
10.1109/
T
I
P
.2021.3066903.
[
22]
S
.
Y
ı
l
dı
z
,
“
T
ur
ki
s
h
s
c
e
ne
t
e
xt
r
e
c
ogni
t
i
on:
i
nt
r
oduc
i
ng
e
xt
e
ns
i
v
e
r
e
a
l
a
nd
s
ynt
he
t
i
c
da
t
a
s
e
t
s
a
nd
a
nove
l
r
e
c
ogni
t
i
on
m
ode
l
,
”
E
ngi
ne
e
r
i
ng Sc
i
e
nc
e
and T
e
c
hnol
ogy
, an I
nt
e
r
nat
i
onal
J
ou
r
nal
, vol
. 60, no. 1,
D
e
c
. 2024, doi
:
10.1016/
j
.j
e
s
t
c
h.2024.101881.
[
23]
Q.
-
D
.
N
guye
n,
N
.
-
M
.
P
ha
n,
P
.
K
r
öm
e
r
,
a
nd
D
.
-
A
.
L
e
,
“
A
n e
f
f
i
c
i
e
nt
uns
upe
r
vi
s
e
d
a
ppr
oa
c
h
f
or
O
C
R
e
r
r
or
c
or
r
e
c
t
i
on
of
V
i
e
t
na
m
e
s
e
O
C
R
t
e
xt
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 11, pp. 58406
–
58421, 2023, doi
:
10.1109/
A
C
C
E
S
S
.2023.3283340.
[
24]
J
.
M
e
m
on,
M
.
S
a
m
i
,
R
.
A
.
K
ha
n,
a
nd
M
.
U
ddi
n,
“
H
a
ndw
r
i
t
t
e
n
opt
i
c
a
l
c
ha
r
a
c
t
e
r
r
e
c
ogni
t
i
on
(
O
C
R
)
:
a
c
om
pr
e
he
ns
i
ve
s
ys
t
e
m
a
t
i
c
l
i
t
e
r
a
t
ur
e
r
e
vi
e
w
(
S
L
R
)
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 8, pp. 142642
–
142668, 2020, doi
:
10.1109/
A
C
C
E
S
S
.2020.3012542.
[
25]
S
.
K
a
r
t
hi
ke
ya
n,
A
.
G
.
S
.
de
H
e
r
r
e
r
a
,
F
.
D
oc
t
or
,
a
nd
A
.
M
i
r
z
a
,
“
A
n
O
C
R
pos
t
-
c
or
r
e
c
t
i
on
a
ppr
oa
c
h
us
i
ng
de
e
p
l
e
a
r
ni
ng
f
o
r
pr
oc
e
s
s
i
ng
m
e
di
c
a
l
r
e
por
t
s
,”
I
E
E
E
T
r
ans
ac
t
i
ons
on C
i
r
c
ui
t
s
and Sy
s
t
e
m
s
f
or
V
i
de
o T
e
c
hnol
ogy
, vol
. 32, no.
5, pp. 2574
–
2581,
M
a
y 2022, doi
:
10.1109/
T
C
S
V
T
.2021.3087641.
B
I
O
G
R
A
P
H
I
E
S
O
F
A
U
T
H
O
R
S
Dr.
Jyoti
Wadmare
is
Assistant
Professor
in
Department
of
Computer
Engineering
at
KJSIT.
She
has
teaching
experience
of
17
years
with
a
n
AI
background.
He
r
major domain
of intere
st
is the conjun
ction of AI
and
computer
visio
n
.
Testimonials of work
includes
many
confere
nces'
presenta
tions
and
articles
published
tha
t
quite
clearly
states
advancement
in th
is area by
her and h
as filed
a patent
and acqui
red fo
ur copyri
ghts.
She ca
n
be contacted at email: jyoti@
somaiya.edu.
Dr.
Sunita
Ravindra
Patil
is
the
Director,
NMIMS
Deemed
to
be
University,
Shirpur
,
Dhule,
Mahar
ashtra
.
She
holds
a
Ph.D.
in
Computer
Engine
ering,
specia
lizing
in
data mining, big data, and
data science
, with around
20 years of
teachi
ng and administrative
experience.
A
member
of
the
board
of
studies
in
computer
engineer
ing
at
UoM,
she
has
published
extensively
in
esteemed
journals
and
conferences
and
has
visited
various
internationa
l
institutions
for
knowledge
exchange
.
Her
focus
is
on
im
plementing
outcome
-
based
academic
reforms
to
benefit
society.
She
can
be
contacted
at
email:
spatil@
somaiya.edu.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 4,
A
ugus
t
2025
:
3412
-
3420
3420
Dakshita K
olte
is a B.Tec
h
.
student in
computer
engineerin
g
at
KJS
IT. She has
developed
AI
and
ML
solutions
that
integrate
artificial
intell
ige
nce
with
web
-
based
technologie
s.
She
has
a
strong
track
record
of
participa
ting
in
prestigi
ous
competitions
such
as Mastek
Project
Deep Blu
e, Aavish
kar, and
Creativ
e Ideas and
Inno
vations
in Acti
on. Sh
e
has
also
been
honored
with
the
prestigious
“Somaiya
star
girl
”
Award
by
Somaiya
Management.
Additionally,
she
holds
four
copyrights
for
her
work.
She
can
be
contacted
at
email:
d.kolt
e@
somaiya.
edu.
Kapil
Bhatia
is
a
B.Tech
.
student
in
computer
en
gineering
at
KJSIT,
specializing
in
artificial
intell
igence
and
machine
learning
.
He
excels
in
deve
loping
solutions
that
integrate
web
-
based
technologies,
the
internet
of
things,
and
artifi
cial
intelligence
.
His
participation
in
renowned
competitions
such
as
Aavishkar,
Mastek
Pr
oject
Deep
Blue,
and
Creative
Ideas
and In
novations
in
Action s
howcas
es his e
x
ceptiona
l ex
pertise
. He hol
ds four
copyrigh
ts
for
his
work
and
has
received
widespread
appreciati
on
for h
is
innovative
project
developments.
He can be contacted at email:
kapil.bhatia@
somaiya
.edu.
Palak
Desai
is
a
computer
engineerin
g
student
at
KJSIT
passi
onate
about
creativit
y
and
technolo
gy
in
the
domains
UI/UX
design,
front
-
end
web
development,
and
data
analytics.
She
enjoys
creating
intuitive,
beautiful
user
interfaces
and
analyzing
data
to
drive
insights.
She
has
participated
in
competitions
like
Aavishkar
and
Creative
Ideas
and
Innovations
in
Action
(CIIA)
for
her
co
ntribution
in
two
institute
-
l
evel
projects.
She
is
dedicated
to
continuous
learning
and
making
impactful,
user
-
centered
soluti
ons.
She
has
been
granted
three
copyrights
for
her
work.
She
can
be
contacted
at
email:
pa
lak.pd@
somaiya.edu.
Ganesh
Wadmare
is
an
Assistant
Profe
ssor
in
the
Department
of
Artificial
Intelligen
ce
and
Data
Science
of
KJSIT
and
a
Ph.D.
scholar
in
S
avitribai
Phule
Pune
University,
with
academic
experience
for
over
19
years.
He
has
ex
tensive
exposure
and
experience
in
the
field
of
artificial
intell
igence
and
renewable
sour
ce
of
energy.
He
has
published
his
research
papers
in
both
national
and
international
co
nferences.
He
can
be
contacted
at email
:
gwadmare@som
aiya.edu.
Evaluation Warning : The document was created with Spire.PDF for Python.