I
AE
S
I
n
t
e
r
n
at
ion
al
Jou
r
n
al
of
Ar
t
if
icial
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
Vol.
14
,
No.
4
,
Augus
t
2025
,
pp.
3143
~
3152
I
S
S
N:
2252
-
8938
,
DO
I
:
10
.
11591/i
jai
.
v
14
.i
4
.
pp
31
43
-
3152
3143
Jou
r
n
al
h
omepage
:
ht
tp:
//
ij
ai
.
iaes
c
or
e
.
c
om
S
u
r
ve
y o
n
3D
b
io
m
e
t
r
ic
t
r
ai
t
s
f
or
h
u
m
a
n
i
d
e
n
t
i
f
ic
at
io
n
Divya
Gan
gac
h
an
n
aiah
1
,
M
am
a
t
h
a
Ar
u
va
n
all
i
S
h
ivaraj
2
,
Hon
gan
u
r
Cha
n
d
r
as
e
k
h
ar
aiah
Naga
r
aj
1
,
P
r
as
an
n
a
Gu
r
u
r
aj
P
aga
1
1
D
e
pa
r
t
me
nt
of
E
l
e
c
tr
o
ni
c
s
a
nd
C
om
muni
c
a
ti
on E
n
gi
ne
e
r
in
g
, N
i
tt
e
M
e
e
na
ks
hi
I
n
s
ti
t
ut
e
of
T
e
c
hno
lo
gy
(
N
M
I
T
)
,
N
itte
(
D
e
e
me
d t
o
be
U
ni
ve
r
s
it
y)
,
B
e
nga
lu
r
u, I
ndi
a
2
D
e
pa
r
tm
e
nt
of
E
le
c
tr
oni
c
s
a
nd C
omm
uni
c
a
ti
on E
ngi
ne
e
r
in
g, N
M
A
M
I
ns
ti
tu
te
of
T
e
c
hnol
ogy
(
N
M
A
M
I
T
)
,
N
itte
(
D
e
e
me
d t
o be
U
ni
ve
r
s
it
y)
,
N
it
te
, I
ndi
a
Ar
t
icle
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
Apr
16,
2024
R
e
vis
e
d
F
e
b
22,
2025
Ac
c
e
pted
M
a
r
15,
2025
In
d
i
v
i
d
u
al
s
are
v
eri
fi
e
d
an
d
i
d
en
t
i
f
i
ed
u
s
i
n
g
Bi
o
met
ri
c
t
ech
n
o
l
o
g
y
b
a
s
ed
o
n
t
h
e
i
r
b
i
o
l
o
g
i
cal
o
r
b
e
h
av
i
o
ra
l
t
ra
i
t
s
.
Bi
o
met
r
i
c
-
b
a
s
ed
p
e
rs
o
n
al
au
t
h
e
n
t
i
cat
i
o
n
s
y
s
t
ems
are
m
o
re
rel
i
ab
l
e
an
d
u
s
e
r
fri
en
d
l
y
,
o
v
erru
n
s
t
h
e
t
rad
i
t
i
o
n
a
l
p
er
s
o
n
al
au
t
h
en
t
i
ca
t
i
o
n
s
y
s
t
ems
.
T
h
e
p
h
y
s
i
o
l
o
g
i
ca
l
b
i
o
me
t
ri
c
t
r
ai
t
s
g
e
t
ab
ra
d
ed
d
u
e
t
o
ag
i
n
g
an
d
mas
s
i
v
e
w
o
rk
,
w
h
i
l
e
t
h
e
b
e
h
av
i
o
ra
l
b
i
o
me
t
ri
c
t
rai
t
s
are
h
av
i
n
g
h
i
g
h
v
ari
a
t
i
o
n
s
d
u
e
t
o
ex
t
ern
a
l
f
act
o
r
s
s
u
c
h
as
fat
i
g
u
e,
an
d
mo
o
d
.
A
mo
n
g
t
h
e
p
h
y
s
i
o
l
o
g
i
ca
l
b
i
o
me
t
ri
c
t
ra
i
t
s
,
Fi
n
g
er
g
e
o
met
r
y
p
a
t
t
ern
s
are
w
i
d
el
y
d
ep
l
o
y
ed
au
t
h
en
t
i
ca
t
i
o
n
s
y
s
t
em
rea
s
o
n
b
e
i
n
g
i
t
s
s
t
a
b
i
l
i
t
y
,
u
s
er
accep
t
a
b
i
l
i
t
y
a
n
d
u
n
i
q
u
en
e
s
s
.
Recen
t
t
ren
d
s
i
n
Bi
o
met
r
i
cs
a
t
t
em
p
t
t
o
i
n
c
o
rp
o
rat
e
3
D
d
o
mai
n
t
rai
t
s
,
3
D
rec
o
n
s
t
r
u
ct
i
o
n
i
s
d
o
n
e
u
s
i
n
g
2
D
mu
l
t
i
p
l
e
i
m
ag
es
.
3
D
i
mag
e
s
are
u
s
u
al
l
y
m
o
re
r
o
b
u
s
t
an
d
i
l
l
u
m
i
n
a
t
i
o
n
i
n
v
ar
i
an
t
as
co
mp
ared
t
o
t
h
ei
r
2
D
co
u
n
t
er
p
art
s
.
3
D
reco
n
s
t
ru
c
t
i
o
n
a
l
g
o
ri
t
h
m
s
are
co
mp
ar
ed
b
y
fi
n
d
i
n
g
mean
s
q
u
are
erro
r
(MSE
).
K
e
y
w
o
r
d
s
:
3D
biom
e
tr
ic
t
r
a
it
B
iom
e
tr
ic
f
inger
knuc
kle
B
iom
e
tr
ic
identif
ier
F
inger
ge
ometr
y
pa
tt
e
r
ns
M
e
a
n
s
qua
r
e
e
r
r
or
P
hys
iol
ogica
l
tr
a
it
s
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
ce
n
s
e.
C
or
r
e
s
pon
din
g
A
u
th
or
:
M
a
matha
Ar
uva
na
ll
i
S
hivar
a
j
As
s
oc
iate
P
r
of
e
s
s
or
,
De
pa
r
tm
e
nt
of
E
lec
tr
onics
a
n
d
C
omm
unica
ti
on
E
nginee
r
ing
NM
AM
I
ns
ti
tut
e
of
T
e
c
hnology
,
N
i
tte
(
De
e
med
to
be
Unive
r
s
it
y)
Nitt
e
,
Ka
r
na
taka
,
I
ndia
E
mail:
mama
thag
__12@r
e
di
f
f
mail.
c
om
1.
I
NT
RODU
C
T
I
ON
T
o
a
c
c
e
s
s
a
va
r
iety
of
r
e
s
our
c
e
s
,
f
r
o
m
c
omput
e
r
s
ys
tems
to
r
e
s
e
a
r
c
h
f
a
c
il
it
ies
a
nd
loca
ti
ons
li
ke
c
oll
e
ge
c
a
mpus
e
s
,
to
nuc
lea
r
plants
,
we
f
r
e
que
ntl
y
ne
e
d
to
a
uthentica
te
our
identit
ies
or
the
identit
ies
of
other
s
.
All
s
e
c
ur
it
y
s
ys
tems
a
r
e
ba
s
e
d
on
a
c
c
e
s
s
c
ontr
ol;
to
dis
c
r
im
inate
be
twe
e
n
law
-
a
bidi
ng
c
it
i
z
e
ns
a
nd
c
r
im
inals
,
the
a
ppr
op
r
iate
indi
v
iduals
s
hould
be
pe
r
mi
tt
e
d
e
nt
r
y
while
the
inappr
op
r
iate
ind
ivi
dua
ls
s
hould
be
ba
r
r
e
d.
Authe
nti
c
a
ti
on
is
the
p
r
oc
e
s
s
of
a
ll
owin
g
only
a
uthentica
ted
indi
viduals
to
a
c
c
e
s
s
the
de
s
ignate
d
pr
otec
ted
r
e
s
our
c
e
s
a
nd
loca
ti
ons
[
1]
,
[
2]
.
T
r
a
dit
io
na
l
a
uthentica
ti
on
s
ys
tems
a
r
e
ba
s
e
d
on
"
W
ha
t
you
know"
(
knowle
dge
-
ba
s
e
d
s
ys
tems
)
,
whic
h
include
s
p
a
s
s
wor
ds
a
nd
pe
r
s
ona
l
identi
f
ica
ti
on
number
s
(
P
I
Ns
)
o
r
a
nythi
ng
e
ls
e
you
c
a
n
r
e
membe
r
a
nd
wr
i
te,
a
nd
"
W
ha
t
you
ha
ve
"
(
token
-
ba
s
e
d
s
ys
tems
)
,
whic
h
i
nc
ludes
phys
ica
l
a
uthentica
ti
on
de
vice
s
.
T
he
be
low
poin
ts
outl
ine
the
pr
im
a
r
y
is
s
ue
s
or
de
f
icie
nc
ies
a
s
s
oc
ia
ted
with
a
uthentica
ti
on
s
ys
tems
ba
s
e
d
on
pa
s
s
wor
ds
,
s
mar
t
c
a
r
ds
,
or
pa
s
s
wor
d
tokens
:
P
a
s
s
wor
ds
a
nd
P
I
Ns
f
unc
ti
on
we
ll
a
s
long
a
s
they
r
e
main
im
pe
r
vious
to
una
ut
hor
ize
d
gue
s
s
ing
a
tt
e
mpt
s
.
T
he
a
c
t
of
s
ha
r
i
ng
pa
s
s
wor
ds
pr
e
s
e
nts
a
s
igni
f
ica
nt
is
s
ue
.
W
hil
e
pa
s
s
wor
d
token
s
or
s
mar
t
c
a
r
ds
a
r
e
not
a
s
e
a
s
il
y
s
ha
r
e
d
a
s
pa
s
s
wor
ds
,
they
a
r
e
none
thele
s
s
s
u
s
c
e
pti
ble
to
thef
t
or
los
s
[
3]
.
T
he
pil
f
e
r
e
d
c
a
r
ds
ha
ve
the
potential
to
be
uti
li
z
e
d
by
a
n
una
uthor
ize
d
indi
vidual
in
or
de
r
to
ob
tain
a
c
c
e
s
s
to
va
r
ious
r
e
s
our
c
e
s
.
T
he
r
e
a
r
e
s
e
ve
r
a
l
vulne
r
a
bil
it
ies
a
nd
c
ha
ll
e
nge
s
t
ha
t
a
r
e
inher
e
ntl
y
li
nke
d
to
wha
t
you
know
a
nd
wha
t
you
ha
ve
a
uthentica
ti
on
s
ys
tems
.
T
he
inher
e
nt
li
mi
tations
of
a
uthentica
ti
on
s
ys
tems
pr
e
ve
nt
them
f
r
om
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
4
,
Augus
t
2025
:
314
3
-
3152
3144
a
c
c
ur
a
tely
dis
c
e
r
ning
the
or
igi
n
of
a
n
a
uthentica
tor
,
na
mely
if
it
o
r
igi
na
tes
f
r
om
a
de
vice
that
ha
s
be
e
n
gue
s
s
e
d,
s
ha
r
e
d,
or
s
tol
e
n
[
4]
.
T
he
s
e
s
ys
tems
pr
ovide
a
s
igni
f
ica
nt
po
tential
f
or
una
utho
r
ize
d
indi
v
iduals
or
c
r
im
inal
e
nti
t
ies
to
r
e
a
dil
y
br
e
a
c
h
the
s
e
c
ur
it
y
mea
s
ur
e
s
a
nd
ge
t
a
c
c
e
s
s
to
the
s
a
f
e
gua
r
de
d
r
e
s
our
c
e
s
.
T
o
e
f
f
e
c
ti
ve
ly
tac
kle
thes
e
c
onc
e
r
ns
,
it
is
im
pe
r
a
ti
ve
to
im
pleme
nt
a
mo
r
e
s
tr
e
a
ml
ined
a
n
d
r
obus
t
a
uthentica
ti
on
s
ys
tem
.
T
he
pr
opos
e
d
s
olut
ion
is
a
n
a
uthentica
ti
on
s
ys
tem
ba
s
e
d
on
b
iom
e
tr
ics
,
s
pe
c
if
ica
ll
y
known
a
s
"
W
ha
t
you
a
r
e
”
[
5]
.
T
he
te
r
m
"
B
iom
e
tr
ic"
is
e
tym
ologi
c
a
ll
y
de
r
ived
f
r
om
the
c
ombi
na
ti
o
n
of
two
Gr
e
e
k
ter
ms
,
na
mely
"
bios
"
whic
h
pe
r
tains
to
li
f
e
,
a
nd
"
metr
os
"
whic
h
de
note
s
mea
s
ur
e
ment.
T
he
pr
oc
e
s
s
of
a
utom
a
ti
c
a
ll
y
ve
r
if
ying
a
nd
identif
ying
a
n
indi
vi
dua
l's
identif
ica
ti
on
is
a
c
c
ompl
is
he
d
by
a
na
lyzin
g
one
or
mor
e
dis
ti
nc
t
phys
iol
og
ica
l
or
be
ha
vior
a
l
f
e
a
tur
e
s
[
6]
–
[
8]
.
T
he
p
r
im
a
r
y
be
ne
f
it
s
o
f
biom
e
tr
ic
s
ys
tems
include
:
t
he
pa
s
s
wor
d,
in
the
c
a
s
e
of
f
o
r
ge
r
y,
or
identit
y
c
a
r
ds
,
in
the
e
ve
nt
of
mi
s
plac
e
ment,
a
r
e
both
e
xa
mpl
e
s
of
it
e
ms
that
a
r
e
not
s
us
c
e
pti
ble
to
f
r
a
udulent
r
e
pli
c
a
ti
on
or
a
c
c
identa
l
los
s
[
9]
.
T
he
dif
f
e
r
e
nti
a
ti
on
be
twe
e
n
a
uthentic
indi
viduals
a
nd
im
pos
ter
s
is
c
on
s
is
ten
tl
y
e
s
tablis
he
d
by
the
dis
c
e
r
nment
of
unique
a
t
tr
ibut
e
s
,
ne
c
e
s
s
it
a
ti
ng
the
a
c
tual
pr
e
s
e
nc
e
of
the
indi
vidual
i
n
que
s
ti
on.
B
iom
e
tr
ic
s
ys
tems
pr
ovide
e
nha
nc
e
d
r
e
li
a
bil
it
y
a
nd
us
e
r
-
f
r
iendlines
s
.
A
biom
e
tr
ic
s
ys
tem
may
be
de
s
c
r
ibed
a
s
a
tec
hnologi
c
a
l
tool
uti
li
z
e
d
to
a
s
c
e
r
tain
the
identit
y
of
a
n
indi
vidual
by
qua
nti
f
ying
one
o
r
ma
ny
phys
iol
ogica
l
or
be
ha
vio
r
a
l
a
tt
r
ibu
tes
.
T
he
ter
m
"
biom
e
tr
ic
modalit
y"
is
us
e
d
to
de
s
c
r
ibe
the
phys
iol
ogi
c
a
l
or
be
ha
vior
a
l
mea
s
ur
e
ments
obtaine
d
f
r
om
a
n
indi
vidual
f
or
the
pu
r
pos
e
of
a
uthentica
ti
on.
T
he
biom
e
tr
ic
modalit
ies
that
a
r
e
wide
ly
a
c
knowle
dge
d
a
nd
a
ppr
ove
d
on
a
global
s
c
a
le
include
f
inger
p
r
int
s
,
pa
l
m
p
r
int
s
,
ha
nd
ge
ometr
y,
f
inger
knuc
kle
pr
int
s
,
f
a
c
i
a
l
r
e
c
ognit
ion,
ir
is
s
c
a
ns
,
r
e
t
inal
s
c
a
ns
,
e
a
r
r
e
c
ognit
ion,
voice
r
e
c
ognit
ion
,
s
ignatur
e
a
na
lys
is
,
a
nd
ke
ys
tr
oke
dyna
mi
c
s
.
T
he
r
e
a
r
e
two
dif
f
e
r
e
nt
c
a
tegor
ies
f
or
biom
e
tr
ic
s
ys
tems
a
s
s
hown
in
F
igur
e
1,
b
e
ha
vior
a
l
a
nd
phys
iol
ogica
l
biom
e
tr
ics
.
T
he
phys
ica
l
qua
li
ti
e
s
of
a
pe
r
s
on
a
r
e
r
e
pr
e
s
e
nted
by
their
phys
iol
ogica
l
biom
e
tr
ic
t
r
a
it
s
,
while
their
be
ha
vior
a
l
biom
e
tr
ic
tr
a
it
s
a
r
e
their
be
ha
vior
a
l
c
ha
r
a
c
ter
is
ti
c
s
[
10]
.
W
hil
e
the
be
ha
vior
a
l
biom
e
t
r
ic
f
e
a
tur
e
s
a
r
e
ve
r
y
v
a
r
iable
due
to
outs
ide
inf
luenc
e
s
li
ke
we
a
r
ines
s
,
m
ood,
a
nd
other
f
a
c
tor
s
,
the
phys
iol
ogica
l
biom
e
t
r
ic
t
r
a
it
s
d
e
ter
ior
a
te
with
a
ge
a
nd
he
a
vy
wor
k
[
11
]
.
F
inger
ge
ometr
y
pa
tt
e
r
ns
a
r
e
one
of
the
phys
iol
ogica
l
b
iom
e
tr
ic
q
ua
li
ti
e
s
that
a
r
e
f
r
e
que
ntl
y
us
e
d
in
a
uthentica
ti
on
s
ys
tems
be
c
a
us
e
of
their
s
tabili
ty,
us
e
r
a
c
c
e
ptanc
e
,
a
nd
uniq
ue
ne
s
s
.
F
inger
knuc
kle
tr
a
it
s
ge
t
wo
r
n
les
s
unde
r
f
inger
ge
ometr
y
pa
tt
e
r
ns
be
c
a
us
e
they
a
r
e
e
xpos
e
d
to
les
s
int
e
ns
e
wor
k.
F
igur
e
1.
B
iom
e
tr
ic
c
las
s
if
ica
ti
on
I
n
the
r
e
a
lm
of
a
uthentica
ti
on
s
ys
tems
,
biom
e
tr
i
c
tr
a
it
s
e
mer
ge
a
s
pa
r
a
mount
,
ha
r
ne
s
s
ing
unique
phys
iol
ogica
l
a
nd
be
ha
vior
a
l
c
ha
r
a
c
ter
is
ti
c
s
to
ide
nti
f
y
indi
v
iduals
with
unpa
r
a
ll
e
led
pr
e
c
is
ion
a
nd
s
e
c
ur
it
y.
T
his
pa
r
a
digm
s
hif
ts
f
r
om
c
onve
nti
ona
l
a
uthent
ica
ti
on
methodologi
e
s
,
s
uc
h
a
s
pa
s
s
wor
ds
a
nd
phys
ica
l
tokens
,
to
biom
e
t
r
ics
,
unde
r
s
c
or
e
s
the
c
r
it
ica
l
vul
ne
r
a
bil
it
ies
na
mely
thef
t
,
los
s
,
a
nd
una
uthor
ize
d
s
ha
r
ing
-
a
s
s
oc
iate
d
with
tr
a
dit
ional
s
ys
tems
.
B
y
leve
r
a
ging
int
r
ins
ic
a
tt
r
ibut
e
s
,
including
but
not
l
i
mi
t
e
d
to
f
inger
pr
int
s
,
ir
is
pa
tt
e
r
ns
,
f
a
c
ial
ge
ometr
y,
a
nd
vo
ice
s
ignatur
e
s
,
biom
e
tr
ic
a
uthentica
ti
on
s
ys
tems
tr
a
ns
c
e
nd
thes
e
li
mi
tations
[
12]
.
T
he
f
inger
knuc
kle
t
r
a
it
r
e
pr
e
s
e
nts
a
n
innovative
b
io
metr
ic
identi
f
ier
,
c
a
pit
a
li
z
ing
on
the
unique
a
nd
s
table
pa
tt
e
r
ns
f
ound
on
the
human
f
inger
's
knuc
kle
r
e
gions
.
T
he
s
e
pa
tt
e
r
ns
,
r
e
s
is
tant
to
we
a
r
a
nd
e
nvir
onmenta
l
c
ha
nge
s
,
of
f
e
r
a
s
e
c
ur
e
a
nd
dis
ti
nc
ti
ve
mea
ns
f
or
human
identif
ica
ti
on
.
Uti
li
z
ing
non
-
i
ntr
us
i
ve
im
a
ging
tec
hnologi
e
s
,
f
inger
knuc
kle
r
e
c
ognit
io
n
s
ys
tems
c
a
ptur
e
a
nd
a
na
lyz
e
thes
e
int
r
ica
te
pa
tt
e
r
ns
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
Ar
ti
f
I
ntell
I
S
S
N:
2252
-
8938
Sur
v
e
y
on
3D
biome
tr
ic
tr
ait
s
for
human
identif
ica
ti
on
(
Div
y
a
Gangac
hannaiah)
3145
pr
ovidi
ng
a
r
e
li
a
ble
method
f
or
ve
r
if
ying
indi
vidua
l
identit
ies
[
13]
.
W
it
h
a
dva
ntage
s
s
uc
h
a
s
high
uniquene
s
s
,
r
e
s
il
ienc
e
to
f
or
ge
r
y,
a
nd
e
a
s
e
of
a
c
quis
it
ion,
the
f
inger
knuc
kle
tr
a
it
is
e
mer
ging
a
s
a
pr
omi
s
ing
s
o
lut
ion
in
the
r
e
a
lm
of
biom
e
t
r
ic
s
e
c
ur
it
y,
f
indi
ng
a
ppli
c
a
ti
ons
in
a
c
c
e
s
s
c
ontr
ol
,
f
or
e
ns
ic
a
na
lys
is
,
a
nd
be
y
ond.
T
he
s
igni
f
ica
nc
e
of
c
ha
r
a
c
ter
is
ti
c
s
in
the
ove
r
a
ll
f
unc
t
ioni
ng
of
a
biom
e
tr
ic
s
ys
tem
c
a
nnot
be
unde
r
s
tate
d.
T
he
c
r
it
e
r
ia
us
e
d
to
a
s
s
e
s
s
the
s
uit
a
bil
it
y
of
phys
ica
l
or
be
ha
vio
r
a
l
tr
a
it
s
of
a
human
a
s
a
biom
e
tr
ic
modali
ty
include
univer
s
a
li
ty,
uniquene
s
s
,
pe
r
mane
nc
e
,
mea
s
ur
a
bil
it
y,
a
c
c
e
ptabili
ty,
pe
r
f
o
r
manc
e
,
a
nd
c
i
r
c
umv
e
nti
on.
2.
M
E
T
HO
D
Us
ing
dis
ti
nc
ti
ve
phys
iol
ogica
l
or
be
ha
vior
a
l
c
ha
r
a
c
ter
is
ti
c
s
,
s
uc
h
a
s
f
inger
pr
int
s
,
f
a
c
ial
f
e
a
tur
e
s
,
ir
is
pa
tt
e
r
ns
,
or
voice
c
ha
r
a
c
ter
i
s
ti
c
s
,
bio
metr
ic
ident
if
ica
ti
on
is
a
mul
ti
-
s
tep
pr
oc
e
s
s
that
c
on
f
ir
ms
a
pe
r
s
on's
identit
y.
T
he
pr
oc
e
dur
e
s
tar
ts
with
e
nr
oll
ment,
in
whic
h
s
pe
c
ialize
d
s
e
ns
or
s
a
r
e
us
e
d
to
c
oll
e
c
t
biom
e
tr
ic
da
ta,
whic
h
is
then
t
r
a
ns
f
or
med
int
o
a
digi
tal
template
a
nd
s
a
f
e
ly
s
tor
e
d
in
a
da
taba
s
e
.
T
he
tec
hnology
ga
ther
s
f
r
e
s
h
biom
e
tr
ic
inf
or
mat
ion
f
r
om
the
pe
r
s
on
dur
ing
identi
f
ica
ti
on
a
nd
a
na
lyze
s
it
to
obtain
unique
c
ha
r
a
c
ter
is
ti
c
s
.
Adva
nc
e
d
matc
hing
tec
hniques
a
r
e
then
us
e
d
to
c
ompar
e
thi
s
p
r
oc
e
s
s
e
d
da
ta
with
the
template
s
tha
t
ha
ve
be
e
n
s
tor
e
d.
2.
1
.
T
yp
e
s
of
b
iom
e
t
r
ic
s
ys
t
e
m
s
ar
e
m
e
n
t
ion
e
d
h
e
r
e
B
iom
e
tr
ic
identif
ica
ti
on
s
ys
tems
ha
ve
be
c
ome
i
nc
r
e
a
s
ingl
y
im
por
tant
in
e
nha
nc
ing
s
e
c
ur
it
y
a
nd
e
ns
ur
ing
a
c
c
ur
a
te
pe
r
s
ona
l
a
uthentica
ti
on
in
va
r
i
ous
domains
.
T
he
s
e
s
ys
tems
uti
li
z
e
a
n
in
divi
dua
l'
s
unique
phys
iol
ogica
l
a
nd
be
ha
vior
a
l
tr
a
it
s
,
thus
pr
ovi
ding
a
r
e
li
a
ble
a
nd
e
f
f
icie
nt
a
lt
e
r
na
ti
ve
to
t
r
a
dit
ional
identif
ica
ti
on
methods
s
uc
h
a
s
pa
s
s
wor
ds
or
identit
y
c
a
r
ds
.
As
tec
hnology
a
dva
nc
e
s
,
va
r
ious
biom
e
tr
ic
modalit
ies
ha
ve
be
e
n
de
ve
loped
-
e
a
c
h
with
dif
f
e
r
e
nt
f
e
a
tur
e
s
,
be
ne
f
i
ts
,
a
nd
li
mi
tations
-
of
f
e
r
ing
a
wi
de
a
r
r
a
y
of
s
olut
ions
tailor
e
d
to
s
pe
c
if
ic
a
ppli
c
a
ti
on
ne
e
ds
.
S
ome
c
omm
only
us
e
d
types
of
biom
e
tr
ic
s
ys
tems
include
:
−
F
inger
pr
int
b
iom
e
tr
ic
s
ys
tem:
t
he
f
inger
pr
int
is
on
e
of
the
mos
t
we
ll
-
known
biom
e
t
r
ic
identif
ie
r
s
a
nd
ha
s
be
e
n
in
us
e
f
or
mor
e
than
a
c
e
ntur
y
due
to
it
s
uniq
ue
ne
s
s
a
nd
dur
a
bil
it
y.
−
Ha
nd
ge
ometr
y
biom
e
tr
ic
s
ys
tem:
t
his
s
im
ple
a
nd
c
os
t
-
e
f
f
e
c
ti
ve
pr
oc
e
dur
e
us
e
s
a
phys
ica
l
ins
pe
c
ti
on
to
c
onf
ir
m
a
pe
r
s
on's
identit
y.
I
nc
luded
a
r
e
ha
nd
s
ize
,
f
inger
length
,
br
e
a
dth,
a
nd
f
or
m
.
z
e
d.
Due
to
i
ts
lac
k
of
dif
f
e
r
e
nc
e
,
thi
s
method's
dr
a
wba
c
k
is
that
it
onl
y
a
ll
ows
f
or
one
-
to
-
one
matc
hing.
B
y
including
ne
w
biom
e
tr
ic
tr
a
it
s
,
ha
nd
r
e
c
ognit
ion
a
c
c
ur
a
c
y
c
a
n
be
incr
e
a
s
e
d.
−
F
a
c
e
r
e
c
ognit
ion
s
ys
te
m:
p
e
ople
may
be
identi
f
i
e
d
a
nd
c
ha
r
a
c
ter
ize
d
by
their
unique
f
a
c
ial
s
tr
uc
tur
e
,
whic
h
c
ons
is
ts
of
pe
a
ks
a
nd
va
ll
e
ys
of
va
r
ied
he
ight
s
.
S
im
ple
ge
ometr
ic
models
ga
ve
wa
y
to
incr
e
a
s
ingl
y
c
ompl
e
x
mathe
matica
l
r
e
pr
e
s
e
ntations
a
nd
matc
hi
ng
a
lgor
it
hms
a
s
f
a
c
e
r
e
c
ognit
ion
tec
hnology
a
dva
nc
e
d.
−
I
r
is
biom
e
tr
ic
r
e
c
ognit
ion
s
ys
tem:
o
ne
o
f
the
mos
t
r
e
li
a
ble
biom
e
t
r
ic
identi
f
ica
ti
on
a
nd
ve
r
if
i
c
a
ti
on
tec
hniques
r
e
c
e
ntl
y
de
ve
loped
is
ir
is
s
c
a
n
tec
hnology.
T
he
biom
e
tr
ic
tec
hnology
s
ys
tem
that
us
e
s
th
e
e
ye
to
identif
y
a
pe
r
s
o
n
is
unde
r
the
e
ye
c
a
tegor
y
.
−
Voic
e
r
e
c
ognit
ion
biom
e
tr
ic
s
ys
tem:
t
he
voice
r
e
c
ognit
ion
s
ys
tem
us
e
s
a
n
indi
vidual’
s
voic
e
f
or
de
ter
mi
ning
identit
y
ba
s
e
d
on
the
di
f
f
e
r
e
nt
c
ha
r
a
c
ter
is
ti
c
voice
f
e
a
tu
r
e
s
.
T
he
s
ys
tem
f
o
r
the
s
ynthes
is
of
the
s
ound
pr
oduc
e
d
by
th
e
lar
ynx
.
2.
2
.
S
t
age
s
in
volved
i
n
a
b
iom
e
t
r
ic
s
ys
t
e
m
d
u
r
i
n
g
r
e
c
ogn
it
ion
B
iom
e
tr
ic
s
ys
tems
unde
r
go
a
s
e
r
ies
of
c
r
uc
ial
s
teps
,
including
p
r
e
-
pr
oc
e
s
s
ing
to
e
nha
nc
e
da
ta
a
c
c
ur
a
c
y,
r
e
gion
of
int
e
r
e
s
t
s
e
lec
ti
on
f
o
r
f
e
a
tur
e
e
xtr
a
c
ti
on
inf
luenc
e
d
by
r
e
tu
r
n
on
inves
tm
e
nt,
a
n
d
f
e
a
tur
e
e
xtr
a
c
ti
on
to
c
ons
tr
uc
t
unique
da
ta
r
e
pr
e
s
e
ntations
.
F
e
a
tur
e
e
xtr
a
c
ti
on
a
im
s
to
r
e
duc
e
da
tas
e
t
s
ize
while
pr
e
s
e
r
ving
e
s
s
e
nti
a
l
inf
or
mation.
M
a
tching
modul
e
s
c
ompar
e
ne
wly
c
r
e
a
ted
template
s
with
r
e
f
e
r
e
nc
e
template
s
,
ge
ne
r
a
ti
ng
matc
h
s
c
or
e
s
to
va
li
da
te
c
l
a
im
e
d
identit
ies
.
Ove
r
a
ll
,
thes
e
s
teps
e
ns
ur
e
e
f
f
ic
ient
a
nd
a
c
c
ur
a
te
biom
e
tr
ic
r
e
c
ognit
ion
p
r
oc
e
s
s
e
s
.
F
igur
e
2
s
hows
the
s
tage
s
invol
ve
d
in
a
b
iom
e
tr
ic
s
ys
tem
dur
ing
r
e
c
ognit
ion
.
−
P
r
e
-
pr
oc
e
s
s
ing:
t
he
biom
e
tr
ic
s
ys
tems
pr
e
-
pr
oc
e
s
s
ing,
f
e
a
tur
e
e
xtr
a
c
ti
on,
matc
hing,
a
nd
de
c
is
ion
-
making
s
teps
a
r
e
include
d.
At
the
pr
e
-
pr
oc
e
s
s
ing
s
tage
,
the
c
omput
e
r
is
e
mpl
oye
d
to
e
nha
nc
e
the
a
c
c
ur
a
c
y
o
f
the
biom
e
tr
ic
da
ta
c
oll
e
c
ted
by
the
s
e
ns
or
.
−
R
e
gion
of
int
e
r
e
s
t
s
tage
:
t
he
r
e
gion
of
int
e
r
e
s
t
a
ppr
oa
c
h
is
us
e
d
by
biom
e
tr
ic
s
ys
tem
s
be
f
or
e
to
or
s
im
ult
a
ne
ous
ly
with
the
f
e
a
tur
e
e
x
t
r
a
c
ti
on
s
tep.
T
he
s
e
lec
ti
on
of
bio
metr
ic
f
e
a
tur
e
qua
li
ti
e
s
to
be
us
e
d
a
s
matc
hing
c
r
it
e
r
ia
in
a
bio
metr
ic
s
ys
tem
is
inf
luenc
e
d
by
r
e
tur
n
on
inves
tm
e
nt.
−
F
e
a
tur
e
e
xtr
a
c
ti
on:
t
he
e
xtr
a
c
ti
on
of
c
ha
r
a
c
ter
is
ti
c
s
is
f
unda
menta
l
f
or
r
e
c
ognit
ion
s
ys
tems
.
T
he
mos
t
e
s
s
e
nti
a
l
inf
or
mation
is
take
n
f
r
om
the
ins
pe
c
ted
biom
e
tr
ic
da
ta
to
c
ons
tr
uc
t
a
ne
w
da
ta
r
e
pr
e
s
e
ntation.
E
ve
r
yone
’
s
idea
l
ne
w
r
e
pr
e
s
e
ntation
would
be
uni
que
.
I
n
the
identif
ica
ti
on
pr
oc
e
s
s
f
or
biom
e
tr
ic
s
ys
tems
,
the
e
xtr
a
c
ti
on
of
c
ha
r
a
c
ter
is
ti
c
s
is
a
n
e
s
s
e
n
ti
a
l
s
ta
ge
.
I
t
invol
ve
s
us
ing
les
s
r
e
s
our
c
e
s
to
a
na
lys
e
s
a
lar
ge
volum
e
of
da
ta
.
B
y
r
e
movi
ng
in
f
or
mation
that
may
be
us
e
d
to
c
a
tegor
ize
a
nd
a
c
quir
e
vis
ua
l
i
nput
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
4
,
Augus
t
2025
:
314
3
-
3152
3146
pa
tt
e
r
ns
,
f
e
a
tu
r
e
e
xt
r
a
c
ti
on's
main
goa
l
is
to
lo
we
r
the
s
ize
of
the
or
ig
inal
da
tas
e
t.
F
or
a
n
o
f
f
li
ne
ha
ndwr
it
ten
s
igni
ng
s
ys
tem,
s
tatic
a
nd
phony
-
dyna
mi
c
f
e
a
tur
e
e
xtr
a
c
ti
on
a
ppr
oa
c
he
s
ha
ve
be
e
n
de
ve
loped.
−
M
a
tching
modul
e
:
t
he
ne
wly
c
r
e
a
ted
template
is
c
ompar
e
d
to
one
or
mor
e
r
e
f
e
r
e
nc
e
template
s
by
the
matc
hing
a
lgor
it
hm.
T
he
r
e
s
ult
of
the
matc
hing
a
lgor
it
hm
is
matc
h
s
c
or
e
,
indi
c
a
ti
ng
how
s
im
il
a
r
the
template
s
a
r
e
.
T
he
number
of
matc
hing
be
twe
e
n
the
input
template
a
nd
the
s
tor
e
d
r
e
f
e
r
e
nc
e
tem
plate
f
e
a
tur
e
s
e
ts
is
de
ter
mi
ne
d
a
nd
a
matc
h
s
c
or
e
r
e
por
ted.
M
a
tch
s
c
or
e
s
a
r
e
us
e
d
to
va
li
da
te
a
c
laimed
ide
nt
it
y
in
or
de
r
to
identi
f
y
a
n
indi
vidual.
F
igur
e
2.
S
tage
s
invol
ve
d
in
a
bio
metr
ic
s
ys
tem
du
r
ing
r
e
c
ognit
ion
2.
2.
1
.
3D
o
n
2D
b
iom
e
t
r
ics
A
unique
mul
ti
-
view
,
mul
ti
-
s
pe
c
tr
a
l
3D
f
inger
i
maging
s
ys
tem
is
the
s
ubjec
t
[
14]
a
s
f
a
r
a
s
we
c
ur
r
e
ntl
y
know,
thi
s
biom
e
tr
ic
im
a
ging
de
vice
s
e
e
ms
to
be
t
he
f
ir
s
t
o
f
it
s
type
that
c
a
n
r
e
c
or
d
a
wid
e
va
r
iety
of
inf
o
r
mation
obtaine
d
f
r
om
the
f
inge
r
.
A
numb
e
r
of
f
inger
s
'
e
xter
na
l
s
kin
a
nd
int
e
r
io
r
ve
ins
we
r
e
im
a
ge
d
us
ing
3D
f
inger
im
a
ging
tec
hnology.
S
ix
dif
f
e
r
e
nt
a
ngles
we
r
e
us
e
d
to
s
c
a
n
the
f
inger
s
in
or
de
r
t
o
a
c
c
ompl
is
h
thi
s
.
W
hil
e
3D
f
inger
knuc
kle
de
tec
ti
on
is
a
n
i
m
a
ge
r
e
c
ognit
ion
c
ha
ll
e
nge
in
a
nd
of
it
s
e
lf
,
Al
-
J
a
na
bi
a
nd
Al
-
J
uboor
i
[
15]
ha
s
de
mons
tr
a
ted
that
s
ubs
tantial
va
r
iation
be
twe
e
n
the
tr
a
in
a
nd
tes
t
da
tas
e
t
dis
tr
ibut
ions
a
nd
a
lac
k
of
tr
a
ini
ng
da
ta
ne
ga
ti
ve
ly
im
pa
c
t
th
e
pe
r
f
or
manc
e
o
f
ge
ne
r
i
c
de
e
p
ne
ur
a
l
ne
twor
ks
,
s
uc
h
a
s
R
e
s
Ne
t.
T
he
s
e
dif
f
iculti
e
s
r
e
s
tr
ict
the
a
bil
it
y
of
g
e
ne
r
ic
ne
ur
a
l
ne
twor
ks
to
c
las
s
if
y
a
uthentic
identi
ti
e
s
in
a
mor
e
ge
ne
r
a
l
wa
y.
T
he
a
dva
nc
e
ment
o
f
3D
r
e
c
ons
tr
uc
ti
on
tec
hniques
ha
s
s
pa
r
ke
d
a
s
ur
ge
in
the
e
xplor
a
ti
on
of
bio
metr
ic
identif
ica
ti
on
thr
ough
the
uti
li
z
a
ti
on
of
3D
da
ta
.
T
he
uti
l
iza
ti
on
of
thr
e
e
-
dim
e
ns
ional
d
a
ta
f
r
om
f
inger
knuc
kle
pa
tt
e
r
ns
of
f
e
r
s
a
ddit
ional
inf
o
r
ma
ti
on
that
is
c
ompl
e
menta
r
y
a
nd
invar
iant
to
il
lu
mi
na
ti
on
[
16]
.
T
his
e
nha
nc
e
s
the
r
e
li
a
bil
i
ty
a
nd
a
c
c
ur
a
c
y
o
f
biom
e
tr
ic
identif
ica
ti
on
.
T
he
uti
li
z
a
ti
on
of
de
e
p
lea
r
ning
tec
hniques
ha
s
unde
r
gone
e
xtens
ive
r
e
s
e
a
r
c
h
f
or
a
wide
r
a
nge
of
c
omput
e
r
v
is
ion
a
ppli
c
a
ti
ons
,
i
nc
ludi
ng
biom
e
tr
ics
.
T
he
uti
li
z
a
ti
on
o
f
de
e
p
lea
r
ning
tec
hniques
wa
s
a
ls
o
e
mpl
oye
d
in
the
c
utt
ing
-
e
dge
de
ve
lopm
e
nt
of
3D
f
inger
knuc
kle
identif
ica
ti
on
.
T
he
pr
opos
e
d
method
a
im
s
to
mi
ti
ga
te
the
is
s
ue
of
une
ve
n
f
inger
knuc
kle
pa
tt
e
r
ns
by
incor
por
a
ti
ng
int
e
r
media
te
inf
or
mation
f
r
om
mul
ti
ple
-
s
c
a
le
de
e
p
ne
ur
a
l
ne
twor
ks
s
im
ult
a
ne
ous
ly.
2.
2.
2
.
3D
r
e
c
on
s
t
r
u
c
t
ion
3D
r
e
c
ons
tr
uc
ti
on
of
biom
e
tr
ic
tr
a
it
s
is
a
n
a
d
va
nc
e
d
tec
hnique
that
invol
ve
s
c
r
e
a
ti
ng
th
r
e
e
-
dim
e
ns
ional
models
of
human
biom
e
tr
ic
f
e
a
tur
e
s
f
or
r
e
c
ognit
ion,
a
uthentica
ti
on
,
a
nd
a
na
lys
is
.
T
his
tec
hnology
e
nha
nc
e
s
s
e
c
ur
it
y,
a
c
c
ur
a
c
y,
a
nd
r
obus
tnes
s
in
biom
e
tr
ic
s
y
s
tems
.
3D
f
inger
pr
int
model
s
c
a
ptur
e
ri
dge
de
pth
a
nd
s
kin
e
las
ti
c
it
y,
of
f
e
r
ing
be
tt
e
r
r
e
s
is
tanc
e
to
s
poof
ing
a
tt
a
c
ks
.
3D
r
e
c
ons
tr
uc
ti
on
is
be
ing
done
us
ing
F
r
a
nkot
-
C
he
ll
a
ppa
(
F
C
)
a
lgor
it
hm
a
nd
P
ois
s
on
s
olver
(
P
O)
a
lgor
i
thm
[
17]
.
T
he
FC
a
lgo
r
it
h
m
is
us
e
d
f
or
int
e
gr
a
ti
ng
a
gr
a
dient
f
ield
int
o
a
s
ur
f
a
c
e
by
e
n
s
ur
ing
int
e
gr
a
bil
it
y
us
ing
F
ou
r
ier
t
r
a
ns
f
or
ms
.
A
PO
is
of
ten
us
e
d
to
s
olve
the
2D
or
3D
P
ois
s
on
e
qua
ti
on
a
nd
is
c
omm
only
e
mpl
oye
d
in
s
c
e
na
r
ios
li
ke
s
ha
pe
-
f
r
om
-
s
ha
ding
or
s
e
a
ml
e
s
s
im
a
ge
c
loni
ng.
2D
m
ult
ipl
e
input
f
or
e
f
inger
f
inger
knuc
kle
im
a
ge
s
of
s
ubje
c
t
1
[
18]
c
ons
ider
ing
s
ix
input
mul
ti
ple
i
mage
s
of
f
inger
kn
uc
kle
pr
int
,
out
of
192
s
ubjec
ts
with
e
a
c
h
s
e
ve
n
im
a
ge
s
of
f
or
e
f
inger
a
nd
s
e
ve
n
im
a
ge
s
of
mi
dd
le
f
inger
.
F
i
gur
e
3
s
hows
the
output
3D
r
e
c
ons
tr
uc
ted
im
a
ge
obtaine
d
f
r
om
FC
a
lgor
it
h
m
whic
h
gives
mor
e
de
pth
inf
or
mation
e
f
f
icie
ntl
y
.
I
n
c
ompar
is
on
with
F
igur
e
4,
de
pth
inf
or
mation
is
dis
tr
ibut
e
d
a
bove
the
r
e
f
e
r
e
nc
e
whi
c
h
im
pli
c
a
tes
nor
maliza
ti
on
is
not
e
f
f
e
c
ti
ve
a
nd
im
a
ge
s
a
r
e
blur
r
e
d
ne
a
r
the
e
dge
s
us
ing
PO
method
.
I
n
3D
im
a
ge
pr
oc
e
s
s
ing
a
nd
a
na
lys
is
,
s
tatis
ti
c
a
l
pa
r
a
mete
r
s
play
a
c
r
uc
ial
r
ole
in
c
ha
r
a
c
ter
izing,
e
va
luating,
a
nd
int
e
r
pr
e
ti
ng
volum
e
tr
ic
da
ta
.
T
he
s
e
pa
r
a
mete
r
s
a
s
s
is
t
in
qua
nti
f
ying
s
ur
f
a
c
e
t
e
xtur
e
s
,
a
s
s
e
s
s
ing
a
lgor
it
hm
pe
r
f
or
manc
e
,
a
nd
modeling
c
ompl
e
x
s
tr
uc
tur
e
s
.
F
igu
r
e
5
gives
a
n
ove
r
view
of
mea
n
s
qua
r
e
e
r
r
or
(
M
S
E
)
plot
,
whic
h
is
wide
ly
us
e
d
metr
ic
f
or
e
va
luating
the
a
c
c
ur
a
c
y
of
3D
im
a
ge
pr
oc
e
s
s
ing
a
lgor
it
hms
.
I
t
qua
nti
f
ies
the
a
ve
r
a
ge
s
qua
r
e
d
dif
f
e
r
e
nc
e
be
twe
e
n
c
or
r
e
s
ponding
e
leme
nts
of
two
3D
da
tas
e
ts
,
typi
c
a
ll
y
a
r
e
c
ons
tr
uc
ted
im
a
ge
a
nd
it
s
gr
ound
tr
ut
h.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
Ar
ti
f
I
ntell
I
S
S
N:
2252
-
8938
Sur
v
e
y
on
3D
biome
tr
ic
tr
ait
s
for
human
identif
ica
ti
on
(
Div
y
a
Gangac
hannaiah)
3147
F
igur
e
3.
3D
i
mage
f
r
om
F
C
a
lgo
r
it
hm
F
igur
e
4.
3D
i
mage
f
r
om
pois
on
s
olver
a
lgo
r
it
hm
F
igur
e
5
.
M
S
E
c
ompar
is
on
be
twe
e
n
FC
a
nd
P
ois
s
on
2.
2.
3
.
Advan
t
age
s
of
3D
ove
r
2D
b
iom
e
t
r
ic
s
ys
t
e
m
s
W
it
h
the
de
ve
lopm
e
nt
o
f
the
global
e
c
onomy
a
nd
inf
o
r
mation
tec
hnology,
pa
r
ti
c
ular
ly
with
the
a
dve
nt
of
the
I
n
ter
ne
t
e
r
a
,
a
gr
owing
number
of
p
r
of
e
s
s
ions
now
r
e
quir
e
r
e
li
a
ble
identit
y
ve
r
if
ica
ti
o
n.
I
n
the
c
ontext
of
da
ta,
identit
y
is
p
r
ogr
e
s
s
ively
digi
ti
z
e
d
a
nd
c
onc
e
a
led
[
19]
.
Ve
r
if
ying
a
pe
r
s
on's
identit
y
a
nd
uphol
ding
inf
or
mation
s
e
c
ur
it
y
a
r
e
c
ha
ll
e
nging
ta
s
ks
in
the
digi
tal
e
r
a
.
B
iom
e
tr
ics
a
r
e
we
ll
-
known
f
ield
of
r
e
s
e
a
r
c
h
due
to
it
s
r
e
li
a
bil
it
y
a
nd
s
im
pli
c
it
y
of
u
s
a
ge
.
T
he
us
e
of
biom
e
tr
ic
r
e
c
ognit
ion
tec
hnolo
gies
a
nd
s
ys
tems
ha
s
gr
e
a
tl
y
be
ne
f
it
e
d
a
number
of
bus
i
ne
s
s
e
s
[
20]
.
Due
to
it
s
a
bil
it
y
to
p
r
ovide
identit
y
ve
r
i
f
ica
ti
on
qua
li
ti
e
s
li
ke
e
a
s
e
,
non
-
r
e
pudiation,
a
nd
f
or
ge
r
y
r
e
s
is
tanc
e
that
tr
a
dit
ional
e
nc
r
ypti
on
c
a
nnot,
biom
e
tr
ic
identif
ica
ti
on
tec
hnology
is
be
c
omi
ng
mor
e
a
nd
mor
e
r
e
leva
nt
in
pe
ople's
da
il
y
li
ve
s
.
Huma
ns
,
on
the
other
ha
nd,
us
e
"
mul
ti
-
biol
ogica
l
f
e
a
tu
r
e
r
e
c
ognit
ion"
to
dis
ti
nguis
h
a
nd
identif
y
pe
ople
ba
s
e
d
on
tr
a
it
s
including
a
ppe
a
r
a
nc
e
(
f
a
c
ial
r
e
c
ognit
ion)
,
s
pe
a
king
s
tyl
e
(
v
oice
r
e
c
ognit
ion)
,
a
nd
s
tr
ide
(
ga
it
r
e
c
ognit
ion
)
.
T
he
f
us
ion
0
20
40
60
80
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
1
11
21
31
41
51
61
71
81
91
1
0
1
1
1
1
1
2
1
1
3
1
1
4
1
1
5
1
1
6
1
1
7
1
1
8
1
1
9
1
2
0
1
2
1
1
2
2
1
2
3
1
2
4
1
2
5
1
M
ea
n
S
qu
a
re
E
rro
r
N
umbe
r of
P
i
x
els
C
ompa
ri
s
on
of
mean
s
qua
re
err
or
O
ri
g
i
na
l
i
ma
g
e
M
S
E
F
C
M
S
E
P
O
M
S
E
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
4
,
Augus
t
2025
:
314
3
-
3152
3148
of
mu
lt
ibi
ologi
c
a
l
f
e
a
tur
e
r
e
c
ognit
ion
ba
s
e
d
on
matc
hing
s
c
or
e
s
is
d
e
ter
mi
ne
d
by
the
f
inal
c
onc
lus
ion,
whic
h
is
ge
ne
r
a
ted
f
r
om
the
matc
hing
s
c
or
e
s
of
thes
e
nu
mer
ous
r
e
c
ognit
ions
[
21]
.
T
his
is
c
r
uc
ial
to
the
a
c
c
ur
a
c
y
of
human
biom
e
tr
ic
ident
if
ica
ti
on
c
a
pa
bil
it
ies
.
S
o,
it
is
c
onc
e
ivable
to
a
s
s
e
r
t
that
mul
ti
biol
ogica
l
identif
ica
ti
on
ba
s
e
d
on
s
im
il
a
r
it
y
s
c
or
e
s
is
e
s
s
e
nti
a
l
to
pe
ople's
c
a
pa
c
it
y
to
e
s
tablis
h
their
own
identit
ies
.
S
o,
a
f
ter
de
mons
tr
a
ti
ng
the
e
s
s
e
nti
a
l
e
xpe
r
ti
s
e
in
r
e
c
ognizing
a
va
r
iety
o
f
biol
ogica
l
f
e
a
tur
e
s
.
B
iom
e
t
r
ic
iden
ti
f
ica
ti
on
a
uthentica
tes
a
n
indi
vidual's
identit
y
by
us
ing
th
e
ir
phys
ica
l
or
be
ha
vior
a
l
c
ha
r
a
c
ter
is
ti
c
s
.
T
he
p
r
oc
e
s
s
of
h
uman
biom
e
tr
ic
identif
ica
ti
on
may
a
ls
o
be
di
vided
int
o
two
s
tage
s
:
"
r
e
gis
tr
a
ti
on"
a
nd
"
r
e
c
o
gnit
ion,
"
whe
r
e
by
the
"
r
e
gis
tr
a
ti
on"
s
tage
invol
ve
s
s
tor
ing
va
r
ious
biom
e
tr
ics
in
memor
y
a
nd
the
"
r
e
c
ognit
io
n"
s
tage
invol
ve
s
r
e
c
a
ll
ing
a
nd
c
ompar
ing
thos
e
biom
e
tr
ic
s.
Applying
pr
e
s
s
ur
e
to
the
c
ha
r
a
c
ter
is
ti
c
be
ing
s
c
a
nne
d
in
tr
a
dit
ional
methods
f
or
a
c
quir
ing
2D
phys
iol
ogica
l
da
ta
r
e
s
ult
s
in
e
las
ti
c
dis
tor
ti
ons
that
ha
ve
a
ne
ga
ti
ve
inf
luenc
e
on
matc
hing
a
c
c
ur
a
c
y.
W
it
h
c
ontac
tl
e
s
s
im
a
ge
s
e
ns
ing,
it
is
pos
s
ibl
e
to
ge
t
c
l
e
a
r
im
a
ge
s
w
hil
e
a
ls
o
pr
e
ve
nti
ng
r
e
s
idual
im
pr
int
s
.
T
r
a
dit
ional
metho
ds
f
or
a
c
quir
ing
c
ontac
tl
e
s
s
2D
im
a
ge
s
a
r
e
p
r
one
to
il
lum
ination
p
r
oblems
a
nd
a
r
e
ope
n
to
s
poof
a
tt
e
m
pts
.
He
nc
e
the
c
r
e
a
ti
on
o
f
c
ontac
tl
e
s
s
3D
im
a
ging
methods
f
or
human
identif
ica
ti
on
a
r
e
r
e
s
is
tant
to
l
ight
f
luctu
a
ti
ons
,
li
mi
ts
pr
ob
lems
with
s
kin
de
f
o
r
mations
,
a
n
d
r
e
duc
e
the
li
ke
li
hood
o
f
s
poof
a
s
s
a
ult
s
.
2.
3.
Uni
m
od
al
b
iom
e
t
r
ic
s
ys
t
e
m
s
A
un
im
o
d
a
l
(
or
s
o
li
tar
y)
bi
om
e
t
r
i
c
s
y
s
t
e
m
u
s
e
s
a
s
in
gl
e
bi
om
e
tr
ic
c
h
a
r
a
c
ter
i
s
t
ic
[
2
2]
or
a
s
i
ng
l
e
i
nf
or
m
a
t
io
n
s
ou
r
c
e
t
o
c
on
f
ir
m
or
i
de
nt
if
y
a
p
e
r
s
o
n.
T
h
e
or
e
ti
c
a
ll
y,
u
ni
mo
da
l
s
y
s
t
e
ms
h
a
v
e
b
e
c
om
e
m
or
e
a
c
c
ur
a
t
e
a
n
d
r
e
l
ia
bl
e
o
ve
r
t
im
e
,
bu
t
i
n
pr
a
c
ti
c
e
,
r
e
g
i
s
tr
a
t
io
n
i
s
s
u
e
s
a
r
is
e
b
e
c
a
u
s
e
o
f
n
on
-
u
ni
ve
r
s
a
l
b
io
m
e
tr
ic
c
h
a
r
a
c
t
e
r
i
s
t
ic
s
,
s
po
of
i
ng,
a
nd
in
a
c
c
ur
a
t
e
d
a
t
a
,
a
s
n
ot
e
d
a
bo
ve
.
F
i
ng
e
r
p
r
i
nt
r
e
c
og
ni
ti
on
[
23]
a
na
ly
s
e
s
th
e
d
i
s
tr
ib
ut
io
n
of
li
ne
s
on
th
e
s
u
r
f
a
c
e
o
f
th
e
f
in
ge
r
us
i
ng
a
me
th
od
ba
s
e
d
o
n
f
i
ng
e
r
p
r
i
nt
r
e
c
og
ni
ti
on
to
l
oo
k
f
or
c
e
r
t
a
i
n
t
r
a
it
s
.
T
h
e
va
s
t
ma
jor
it
y
of
c
o
n
s
u
m
e
r
s
a
r
e
op
e
n
t
o
u
s
in
g
f
in
g
e
r
pr
in
t
id
e
n
ti
f
ic
a
t
io
n
a
s
a
ty
p
e
of
b
io
me
tr
i
c
s
e
c
ur
it
y,
e
v
e
n
if
a
s
i
z
e
a
bl
e
por
ti
on
of
th
e
g
e
ne
r
a
l
p
ub
li
c
i
s
u
s
e
d
to
it
.
T
h
e
t
e
c
hn
ol
ogy
i
s
a
l
s
o
f
un
c
t
io
na
l
a
nd
a
c
c
e
s
s
ib
le.
I
t
i
s
im
por
ta
nt
to
k
e
e
p
in
mi
nd
th
a
t
di
f
f
e
r
e
nt
f
in
ge
r
pr
in
t
r
e
c
og
nit
io
n
s
y
s
t
e
m
s
h
a
ve
va
r
yi
ng
a
c
c
e
pt
a
n
c
e
a
nd
r
e
j
e
c
ti
on
e
r
r
or
r
a
t
e
s
.
W
hi
l
e
c
omp
a
r
in
g
t
wo
f
a
c
e
s
,
a
f
a
c
i
a
l
r
e
c
o
gn
it
ion
s
y
s
t
e
m
[
24]
e
x
a
mi
n
e
s
ho
w
va
r
io
u
s
f
a
c
i
a
l
f
e
a
tur
e
s
a
r
e
a
r
r
a
n
g
e
d.
S
o
me
ti
me
s
,
e
xt
e
r
na
l
c
h
a
r
a
c
t
e
r
i
s
ti
c
s
l
ik
e
t
h
e
s
k
in
a
r
e
a
l
s
o
c
o
n
s
i
de
r
e
d.
B
io
me
tr
i
c
f
a
c
e
r
e
c
o
gn
it
i
on
i
s
p
o
s
s
i
bl
e
by
f
a
c
e
d
e
t
e
c
t
io
n
te
c
h
no
lo
gy,
wh
i
c
h
c
a
n
r
e
c
o
gni
z
e
s
e
v
e
r
a
l
f
a
c
e
s
in
p
ic
tur
e
s
.
I
f
a
r
e
mo
te
r
e
c
o
gn
it
i
on
s
y
s
t
e
m
i
s
r
e
q
uir
e
d,
r
e
c
e
nt
a
d
v
a
n
c
e
me
nts
in
thi
s
t
e
c
h
no
lo
gy
ma
k
e
it
a
gr
e
a
t
c
h
oi
c
e
f
or
bi
om
e
tr
ic
s
e
c
ur
it
y.
M
or
e
ov
e
r
,
t
h
e
t
e
c
h
nol
og
y'
s
c
a
p
a
bi
li
ty
to
"
n
e
ga
ti
ve
ly
id
e
n
ti
f
y
"
p
e
o
pl
e
or
r
e
m
ov
e
f
a
c
e
s
m
a
k
e
s
it
m
u
c
h
e
a
s
i
e
r
t
o
s
p
ot
s
u
s
p
ic
io
u
s
p
e
o
pl
e
i
n
a
c
r
o
wd.
A
s
c
a
nn
e
r
l
oo
k
s
a
t
t
he
d
i
s
ti
n
c
ti
v
e
c
h
a
r
a
c
t
e
r
i
s
t
i
c
s
o
f
th
e
ir
i
s
d
ur
i
ng
a
n
ir
i
s
s
c
a
n
[
25]
,
wh
ic
h
le
a
ds
in
t
he
r
e
c
or
di
ng
of
t
ho
s
e
c
h
a
r
a
c
te
r
i
s
t
ic
s
a
s
a
(
ba
r
)
c
od
e
.
P
a
r
ti
c
u
lar
ly
wh
e
n
d
on
e
wit
h
i
nf
r
a
r
e
d
li
g
ht,
ir
i
s
s
c
a
nn
in
g
i
s
r
e
c
o
gn
i
z
e
d
a
s
a
r
e
l
i
a
bl
e
bi
om
e
tr
ic
s
e
c
ur
it
y
t
e
c
h
ni
qu
e
.
P
a
l
m
v
e
i
n
p
a
tt
e
r
n
r
e
c
o
gn
it
io
n
i
s
b
a
s
e
d
on
r
e
c
o
gni
z
in
g
uni
qu
e
v
e
i
n
pa
tt
e
r
n
s
[
26]
.
W
h
il
e
u
s
i
ng
m
or
e
r
e
f
e
r
e
n
c
e
po
int
s
t
ha
n
f
i
ng
e
r
v
e
in
pa
tt
e
r
n
r
e
c
o
gni
ti
on,
t
hi
s
m
e
t
ho
d
of
id
e
n
ti
f
yi
ng
i
s
e
a
s
i
e
r
a
n
d
mor
e
s
e
c
ur
e
.
A
d
a
pt
iv
e
li
n
e
a
r
i
nt
e
r
pol
a
t
io
n
a
l
go
r
it
hm
i
s
u
s
e
d
f
or
3
D
r
e
c
on
s
t
r
u
c
t
io
n
[
27]
,
wh
ic
h
u
s
e
s
tr
i
a
ng
ul
a
ti
on
m
e
th
od.
T
h
e
m
os
t
r
e
f
in
e
d
bi
om
e
tr
ic
s
e
c
ur
i
ty
s
y
s
t
e
m
c
ur
r
e
ntl
y
i
n
u
s
e
u
s
e
s
n
on
r
e
pl
ic
a
b
l
e
ir
i
s
s
c
a
nn
in
g
t
e
c
h
nol
og
y
(
or
c
a
n
o
nly
b
e
r
e
c
r
e
a
t
e
d
wit
h
g
r
e
a
t
d
if
f
ic
ul
ty)
.
A
h
ig
h
le
ve
l
of
c
on
ve
ni
e
n
c
e
i
s
of
f
e
r
e
d
v
ia
q
ui
c
k
a
n
d
a
c
c
ur
a
t
e
pa
lm
s
c
a
n
ni
ng
f
o
r
th
e
u
s
e
r
.
M
o
r
e
ov
e
r
,
u
nim
od
a
l
bio
m
e
tr
i
c
t
e
c
hn
ol
og
ie
s
a
r
e
le
s
s
u
s
e
f
ul
t
ha
n
th
e
y
wo
ul
d
be
i
n
pr
a
c
ti
c
a
l
s
e
t
ti
ngs
.
T
h
us
,
o
n
e
s
ol
uti
on
to
t
he
s
e
i
s
s
u
e
s
i
s
a
m
ul
ti
m
od
a
l
bi
ome
tr
ic
i
d
e
nt
if
i
c
a
t
io
n
s
y
s
t
e
m.
A
dd
it
io
na
l
r
e
s
e
a
r
c
h
i
s
n
e
e
d
e
d
o
n
t
h
e
dr
a
wb
a
c
k
s
o
f
u
nim
od
a
l
bi
om
e
tr
i
c
s
y
s
t
e
m
s
.
−
Nois
y
da
ta
biom
e
tr
ic
da
ta
oc
c
a
s
ionally
c
ontains
no
is
e
whe
n
s
e
ns
or
s
a
r
e
not
pr
ope
r
ly
maintaine
d.
−
Non
-
u
niver
s
a
li
ty
biom
e
tr
ic
s
ys
tem
is
ter
med
un
iv
e
r
s
a
l
whe
n
a
ll
us
e
r
s
c
a
n
identi
f
y
thems
e
lves
us
ing
the
s
a
me
biom
e
tr
ic
f
e
a
tur
e
.
Ar
o
und
2%
o
f
the
popul
a
ti
on,
including
thos
e
with
dis
a
bil
it
ies
a
nd
other
s
who
r
un
a
c
r
os
s
numer
ous
r
oa
dblocks
dur
ing
a
r
e
gular
r
e
gis
tr
a
ti
on
pr
oc
e
s
s
,
a
r
e
known
to
be
una
ble
to
p
r
o
duc
e
high
-
qua
li
ty
f
inger
pr
int
s
.
−
L
a
c
k
of
indi
vidualit
y
s
im
il
a
r
f
e
a
tu
r
e
s
may
be
ga
th
e
r
e
d
us
ing
biom
e
t
r
ic
tec
hnology
,
s
uc
h
a
s
a
f
a
c
e
r
e
c
ognit
ion
s
ys
tem
that
take
s
pictur
e
s
o
f
the
f
a
c
e
.
S
o
me
s
it
ua
ti
ons
include
f
a
ther
a
nd
s
on
o
r
ide
nti
c
a
l
twins
.
M
or
e
f
a
ls
e
matc
he
s
ha
ppe
n
be
c
a
us
e
of
the
u
niquene
s
s
pr
oblem.
A
mul
ti
modal
ult
r
a
s
onic
r
e
c
ognit
ion
s
ys
tem
is
e
xp
e
r
im
e
ntally
e
va
luate
d
on
the
ba
s
is
of
the
f
us
ion
of
3D
ha
nd
ge
ometr
y
a
nd
3D
pa
lm
pr
int
da
ta.
T
he
tec
hnique
pr
oduc
e
s
a
volum
e
t
r
ic
im
a
ge
of
the
c
ompl
e
te
ha
nd
a
nd
divi
de
s
it
int
o
s
e
ve
r
a
l
two
-
dim
e
n
s
ional
pictu
r
e
s
with
dif
f
e
r
e
nt
de
pths
f
or
e
a
c
h
c
ha
r
a
c
ter
is
ti
c
.
T
he
2D
pr
ope
r
ti
e
s
of
e
a
c
h
im
a
ge
a
r
e
ga
ther
e
d
a
nd
then
a
ppr
opr
iate
ly
c
ombi
ne
d
in
or
de
r
to
c
r
e
a
te
a
3D
te
mpl
a
te.
a
gr
ound
-
br
e
a
king
biom
e
tr
ic
method
us
ing
non
-
c
ontac
t
3D
f
inger
na
il
s
c
a
nning.
W
it
h
thi
s
method
,
f
inger
knuc
kles
a
r
e
s
im
ult
a
ne
ous
ly
pho
togr
a
phe
d
in
3D
a
nd
2D,
whic
h
a
ll
ows
f
or
a
leve
l
of
pr
e
c
is
ion
in
matc
hing
that
may
not
be
pos
s
ibl
e
with
only
2D
or
3D
pa
tt
e
r
ns
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
Ar
ti
f
I
ntell
I
S
S
N:
2252
-
8938
Sur
v
e
y
on
3D
biome
tr
ic
tr
ait
s
for
human
identif
ica
ti
on
(
Div
y
a
Gangac
hannaiah)
3149
2.
4
.
M
u
lt
im
od
al
b
iom
e
t
r
ic
s
ys
t
e
m
s
T
hi
s
me
th
od
i
nt
e
gr
a
t
e
s
th
e
r
e
s
ult
s
of
s
e
v
e
r
a
l
bi
om
e
tr
ic
tr
a
it
s
id
e
n
ti
f
ic
a
t
io
n.
A
mu
lt
i
mo
da
l
b
io
me
tr
i
c
s
y
s
t
e
m
e
m
pl
oy
s
s
e
ve
r
a
l
b
io
me
tr
i
c
m
od
a
li
t
ie
s
to
pr
ov
id
e
a
h
ig
hl
y
p
r
e
c
i
s
e
a
n
d
s
e
c
ur
e
bi
om
e
t
r
i
c
i
d
e
nt
if
i
c
a
t
io
n
s
y
s
t
e
m,
i
n
c
o
ntr
a
s
t
to
a
u
ni
mo
da
l
b
io
me
tr
i
c
s
y
s
te
m,
wh
ic
h
mi
gh
t
le
a
d
to
n
on
-
un
ive
r
s
a
li
ty
[
28]
.
O
ne
e
l
e
me
nt
th
a
t
c
omm
on
ly
r
e
s
ult
s
in
i
na
c
c
ur
a
c
i
e
s
i
s
t
h
e
d
e
t
e
r
ior
a
t
io
n
o
f
f
i
ng
e
r
pr
i
nt
s
.
An
e
r
r
or
o
r
f
a
i
lur
e
in
a
m
ult
im
od
a
l
bi
om
e
t
r
i
c
s
y
s
t
e
m
ma
y
no
t
ha
v
e
an
im
p
a
c
t
o
n
a
p
e
r
s
o
n
du
e
to
th
e
a
va
il
a
b
il
it
y
o
f
m
a
ny
bi
om
e
tr
ic
t
e
c
hn
ol
og
y
s
y
s
te
m
s
.
T
h
u
s
,
a
mu
lt
im
od
a
l
s
y
s
t
e
m
'
s
p
ot
e
n
ti
a
l
to
lo
w
e
r
t
h
e
n
on
-
e
n
r
o
lm
e
n
t
r
a
t
e
i
s
on
e
of
it
s
ma
in
b
e
n
e
f
it
s
.
B
y
s
t
a
nd
a
r
d
i
z
i
ng
a
nd
b
a
l
a
nc
in
g
th
e
ge
om
e
tr
ic
m
e
a
n
a
nd
h
yp
e
r
bo
li
c
t
a
ng
e
n
t,
f
a
c
e
a
nd
vo
ic
e
w
e
r
e
br
ou
gh
t
t
og
e
th
e
r
.
R
o
s
s
a
nd
J
a
in
u
s
e
d
li
n
e
a
r
di
s
c
r
i
mi
na
nt
-
b
a
s
e
d
a
lg
or
it
h
m
s
,
t
h
e
s
u
m
r
ul
e
,
a
n
d
a
d
e
c
i
s
i
on
tr
e
e
t
o
m
e
r
g
e
f
a
c
e
,
f
in
ge
r
pr
i
nt,
a
nd
h
a
n
d
g
e
om
e
tr
y
bio
m
e
tr
i
c
s
.
T
h
e
s
u
m
r
ul
e
o
ut
p
e
r
f
o
r
m
e
d
t
h
e
o
th
e
r
s
,
a
c
c
o
r
di
ng
to
th
e
a
ut
hor
s
'
f
in
di
ng
s
.
Ap
pr
oa
c
h
e
s
f
or
i
nt
e
gr
a
ti
n
g
s
p
e
e
c
h
a
n
d
f
a
c
i
a
l
b
io
me
tr
i
c
s
[
29
]
l
oo
ke
d
a
t
mu
lt
il
a
y
e
r
pe
r
c
e
pti
o
n,
s
up
por
t
ve
c
t
or
ma
c
h
in
e
s
,
a
n
d
tr
e
e
c
l
a
s
s
if
ier
s
.
A
m
ul
ti
m
od
a
l
bi
om
e
tr
ic
s
y
s
t
e
m
i
s
c
r
e
a
te
d
u
s
i
ng
r
id
g
e
-
b
a
s
e
d
f
i
n
ge
r
pr
i
nt
m
a
tc
hi
ng
a
n
d
E
ig
e
n
f
a
c
e
m
a
t
c
hi
ng.
T
he
bi
om
e
tr
ic
s
e
n
s
or
n
e
e
d
s
to
be
li
nk
e
d
to
t
h
e
p
r
o
p
e
r
u
s
e
r
i
nt
e
r
f
a
c
e
be
f
or
e
t
he
f
ir
s
t
mo
dul
e
,
the
s
e
n
s
or
m
od
ul
e
,
m
a
y
c
oll
e
c
t
th
e
u
s
e
r
'
s
un
pr
oc
e
s
s
e
d
b
io
me
tr
ic
da
t
a
.
T
h
e
r
a
w
bio
m
e
tr
i
c
da
ta
th
a
t
w
a
s
c
a
p
tur
e
d
a
n
d
s
e
n
t
w
a
s
th
e
n
ut
il
i
z
e
d
to
e
xt
r
a
c
t
c
h
a
r
a
c
t
e
r
i
s
ti
c
s
.
T
h
e
b
io
me
tr
i
c
d
a
t
a
c
o
ll
e
c
te
d
u
s
in
g
th
i
s
a
ppr
oa
c
h
i
s
of
a
d
e
q
u
a
t
e
qu
a
l
it
y
f
or
f
ur
t
he
r
pr
o
c
e
s
s
i
ng.
T
o
c
om
p
a
r
e
th
e
qu
a
li
t
y
,
th
e
a
tt
r
ib
ut
e
s
a
r
e
t
ur
ne
d
in
to
a
di
git
a
l
r
e
pr
e
s
e
n
ta
ti
on
a
n
d
d
e
l
iv
e
r
e
d
to
th
e
m
a
t
c
hin
g
m
odu
l
e
.
T
he
f
inger
pr
int
is
one
o
f
the
mos
t
we
ll
-
known
bi
ometr
ic
identi
f
ier
s
a
nd
ha
s
be
e
n
in
us
e
f
o
r
mor
e
than
a
c
e
ntur
y
due
to
it
s
uniquene
s
s
a
nd
du
r
a
bil
it
y.
F
e
a
tu
r
e
e
xtr
a
c
ti
on
us
ing
3D
ge
ometr
y
of
s
ur
f
a
c
e
nor
mal
ve
c
tor
s
,
f
or
a
c
c
ur
a
tely
e
nc
oding
the
c
ur
va
tur
e
inf
o
r
mation
[
30
]
.
F
e
a
tur
e
c
ompar
is
on
us
ing
s
im
il
a
r
it
y
f
unc
ti
on
ge
ne
r
a
ted
f
r
om
s
tatis
ti
c
a
l
d
is
tr
ibut
ion
of
the
e
n
c
ode
d
f
e
a
tur
e
s
pa
c
e
,
di
f
f
icult
to
de
s
ign
c
ompl
e
x
f
e
a
tur
e
de
s
c
r
ipt
or
with
f
inger
de
f
or
mations
I
ts
br
oa
d
us
e
a
nd
long
-
ter
m
c
oll
e
c
ti
on
by
im
mi
gr
a
ti
on
a
nd
law
e
nf
or
c
e
ment,
a
s
we
ll
a
s
it
s
numer
ous
da
ta
c
oll
e
c
ti
on
s
our
c
e
s
,
including
the
ten
f
inger
s
,
ha
ve
c
ontr
ibut
e
d
to
i
ts
e
nor
mous
popular
it
y
.
T
his
s
im
ple
a
nd
c
os
t
-
e
f
f
e
c
ti
ve
pr
oc
e
dur
e
us
e
s
a
phys
ica
l
ins
pe
c
ti
on
to
c
onf
ir
m
a
pe
r
s
on's
identit
y.
I
nc
luded
a
r
e
ha
nd
s
ize
,
f
inger
le
ngth,
b
r
e
a
dth,
a
nd
f
or
m
.
B
e
c
a
us
e
to
it
s
a
da
ptabili
t
y,
s
oc
ial
a
c
c
e
ptabili
ty,
a
nd
int
e
gr
a
ti
on
potenti
a
l
,
thi
s
biom
e
tr
ic
a
ppr
oa
c
h
is
e
xtens
ively
uti
li
z
e
d
[
31]
,
[
32
]
.
Due
to
it
s
lac
k
of
dif
f
e
r
e
nc
e
,
thi
s
method's
dr
a
wba
c
k
is
that
it
only
a
ll
ows
f
or
one
-
to
-
one
matc
hing.
B
y
includ
ing
ne
w
biom
e
tr
ic
tr
a
it
s
,
ha
nd
r
e
c
ognit
ion
a
c
c
ur
a
c
y
c
a
n
be
incr
e
a
s
e
d.
P
e
ople
may
be
identif
ied
a
nd
c
ha
r
a
c
ter
ize
d
by
their
unique
f
a
c
ial
s
tr
uc
tur
e
,
whic
h
c
ons
is
ts
of
pe
a
ks
a
nd
va
ll
e
ys
of
va
r
ied
he
ight
s
.
T
he
biom
e
tr
ic
s
ys
tem
us
e
s
thi
s
tr
a
it
to
di
f
f
e
r
e
nti
a
te
be
twe
e
n
pe
ople.
T
he
f
a
c
e
s
c
a
n
r
e
c
or
ds
a
nd
s
tor
e
s
a
pe
r
s
on's
f
a
c
e
f
or
f
utur
e
e
nr
olm
e
nt
ve
r
if
ica
ti
on
pur
pos
e
s
.
S
im
ple
ge
ometr
ic
mod
e
ls
ga
ve
wa
y
to
incr
e
a
s
ingl
y
c
ompl
e
x
mathe
matica
l
r
e
pr
e
s
e
ntations
a
nd
matc
hing
a
lgor
it
h
ms
a
s
f
a
c
e
r
e
c
ognit
ion
tec
hnology
a
dva
nc
e
d.
One
of
the
mos
t
r
e
li
a
ble
biom
e
tr
ic
identif
ica
ti
on
a
nd
ve
r
i
f
ica
ti
on
tec
hniq
ue
s
r
e
c
e
ntl
y
de
ve
loped
is
ir
is
s
c
a
n
te
c
hnology
[
17]
.
T
he
biom
e
tr
ic
tec
hnology
s
ys
tem
that
us
e
s
the
e
ye
to
i
de
nti
f
y
a
pe
r
s
on
is
unde
r
the
e
ye
c
a
tegor
y.
T
he
m
e
thod
of
identif
ying
s
omeone
ba
s
e
d
on
their
ir
is
pa
tt
e
r
n
is
known
a
s
ir
is
r
e
c
ognit
ion
.
A
pe
r
s
on
ne
e
ds
a
wr
it
ing
ins
tr
ument
in
their
ha
nd
in
or
de
r
to
s
ign
us
ing
the
ha
ndwr
it
ing
a
nd
s
ignatur
e
tec
hnology
that
h
a
s
be
e
n
a
ppr
ove
d
by
the
gove
r
nment,
the
lega
l
s
ys
tem,
a
nd
c
ompanie
s
a
nd
is
uti
li
z
e
d
by
the
major
i
ty
o
f
indus
tr
ies
f
o
r
identif
ica
ti
on
a
nd
ve
r
if
ica
ti
on
.
Us
ing
dyna
mi
c
s
ig
na
tur
e
r
e
c
ognit
ion
a
nd
a
pe
r
s
on's
tr
a
it
s
,
a
loca
ti
on
may
be
de
ter
mi
ne
d.
T
o
do
th
is
,
it
is
ne
c
e
s
s
a
r
y
to
a
na
lyze
the
X,
Y,
a
nd
Z
a
xe
s
'
dis
ti
nc
ti
ve
s
tr
oke
s
'
s
pe
e
d,
ve
locity,
ti
mi
ng,
a
nd
di
r
e
c
ti
on
[
22
]
.
C
ontac
tl
e
s
s
3D
im
a
ge
s
f
or
identif
ica
ti
on
a
r
e
invar
iant
to
the
c
h
a
nge
s
of
il
lum
inations
,
pos
e
s
les
s
pr
oblems
r
e
late
d
to
the
de
f
or
mations
of
the
s
kin
a
nd
s
poof
a
tt
a
c
ks
.
I
mage
s
e
gmenta
ti
on
us
ing
de
e
p
lea
r
ning
method
us
ing
m
a
s
k
R
-
C
NN
[
3
3
]
e
xtr
a
c
ts
the
dis
c
r
im
inati
ve
f
e
a
tur
e
s
of
f
inger
knuc
kle,
p
a
lm
,
a
nd
f
a
c
e
f
r
om
the
3D
s
ur
f
a
c
e
nor
mal
ve
c
tor
s
by
us
ing
s
ur
f
a
c
e
g
r
a
dient
de
r
ivatives
.
M
a
tching
a
ppr
oa
c
h
us
e
s
s
ur
f
a
c
e
ke
y
point
s
f
o
r
e
s
ti
mating
the
f
inal
s
hif
ti
ng
pa
r
a
mete
r
s
,
whic
h
r
e
d
uc
e
s
the
c
omput
a
ti
ona
l
c
ompl
e
xit
y.
How
e
ve
r
dif
f
icult
to
d
e
s
ign
c
ompl
e
x
f
e
a
tur
e
de
s
c
r
ipt
or
with
f
inger
de
f
or
mations
.
C
NN
a
ppr
oa
c
h
f
or
r
e
c
ognit
ion
f
r
om
de
e
ply
lea
r
nt
mul
ti
s
c
a
le
f
e
a
tur
e
s
a
nd
a
li
gnment
model
r
e
quir
e
s
a
lot
of
tr
a
ini
ng
da
ta
[3
4
]
.
B
iom
e
tr
ic
s
e
ns
ing
s
ys
tems
ha
ve
be
c
ome
mor
e
c
omm
on,
be
c
a
us
e
of
it
s
a
bil
it
y
to
make
us
e
of
dis
ti
nc
ti
ve
biol
ogica
l
qua
li
ti
e
s
.
T
he
gove
r
nment,
a
s
we
ll
a
s
pr
ivate
a
nd
publi
c
or
ga
niza
ti
on
s
,
may
us
e
tec
hnol
ogy
to
f
ight
f
r
a
ud
a
nd
identit
y
thef
t
.
B
io
metr
ic
s
e
ns
ing
tec
hnology
us
e
d
to
be
the
s
a
f
e
s
t
method
of
identif
ying
a
nd
va
li
da
ti
ng
pe
ople,
a
im
s
f
or
publi
c
s
a
f
e
ty.
De
s
pit
e
thi
s
,
high
-
dim
e
ns
ional
da
t
a
with
plenty
of
r
e
dunda
nt
a
nd
unc
or
r
e
late
d
c
ha
r
a
c
ter
is
ti
c
s
s
ti
ll
pr
e
s
e
nt
c
omput
a
ti
ona
l
c
ompl
e
xit
y
is
s
ue
s
.
I
n
o
r
de
r
t
o
r
e
duc
e
dim
e
ns
ionalit
y,
s
pe
e
d
up
c
omput
a
ti
ons
,
a
nd
incr
e
a
s
e
pr
e
c
is
ion,
a
s
ubs
e
t
of
pe
r
ti
ne
nt
c
ha
r
a
c
ter
is
ti
c
s
is
c
hos
e
n.
T
he
r
e
a
r
e
both
unim
oda
l
a
nd
mul
ti
modal
biom
e
t
r
i
c
tec
hnologi
e
s
.
T
he
2D
or
3D
unim
o
da
l
biom
e
tr
ic
s
ys
tem
goa
l
is
to
identif
y
pe
ople
us
ing
jus
t
one
biom
e
tr
ic
tr
a
it
.
T
his
method
f
a
ll
s
s
hor
t
a
nd
is
una
ble
to
pr
ovide
e
nough
r
e
c
ognit
ion
a
c
c
ur
a
c
y.
Da
ta
f
r
om
s
e
ve
r
a
l
b
iom
e
tr
ic
tr
a
it
s
a
r
e
include
d
int
o
the
mul
ti
modal
bi
ometr
ic
s
ys
tem.
I
t
is
mor
e
s
e
c
ur
e
than
a
unim
oda
l
s
ys
tem
a
nd
ha
s
the
a
bil
it
y
to
ge
t
be
yond
is
s
ue
s
li
ke
e
r
r
a
ti
c
s
e
ns
or
da
ta,
lac
k
of
univer
s
a
li
ty,
uniquene
s
s
,
a
nd
biom
e
tr
i
c
tr
a
it
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
4
,
Augus
t
2025
:
314
3
-
3152
3150
3.
CONC
L
USI
ON
I
n
the
c
ompar
is
on
be
twe
e
n
3D
a
nd
2D
biom
e
t
r
ic
identif
ica
ti
on
tec
hnologi
e
s
,
the
thr
e
e
-
dim
e
ns
i
ona
l
a
ppr
oa
c
h
e
xc
e
ls
in
c
a
ptur
ing
the
int
r
ins
ic
de
tails
a
nd
in
-
de
pth
inf
o
r
mation
o
f
biom
e
tr
ic
tr
a
it
s
,
of
f
e
r
in
g
a
mo
r
e
a
c
c
ur
a
te,
s
e
c
ur
e
,
a
nd
r
e
li
a
ble
mea
ns
of
identif
ica
ti
on.
I
ts
r
obus
tnes
s
to
e
nvir
onmenta
l
va
r
iati
ons
a
nd
e
nha
nc
e
d
a
nti
-
s
poof
ing
c
a
pa
bil
it
ies
f
ur
ther
s
oli
dif
y
it
s
a
dva
ntage
ove
r
2D
im
a
ging,
whic
h
is
mor
e
s
u
s
c
e
pti
ble
to
f
o
r
ge
r
y
a
nd
les
s
e
f
f
e
c
ti
ve
unde
r
va
r
y
ing
c
ondit
ions
.
T
he
us
e
r
-
f
r
iendly,
c
ontac
tl
e
s
s
na
tur
e
of
3D
im
a
ging
tec
hnology
not
only
im
pr
ove
s
the
us
e
r
e
xpe
r
ienc
e
but
a
ls
o
e
nc
our
a
ge
s
wide
r
a
dopti
on
a
c
r
os
s
va
r
ious
s
e
c
ur
it
y
-
s
e
ns
it
ive
a
ppli
c
a
ti
ons
.
3D
r
e
c
ons
tr
uc
ti
on
a
lgor
it
hms
a
r
e
be
ing
a
na
lyze
d
by
f
indi
ng
M
S
E
.
M
S
E
is
c
los
e
to
z
e
r
o
in
F
C
a
lgo
r
it
hm
c
ompar
e
d
to
PO
a
lg
or
it
hm.
T
hus
,
whe
n
i
t
c
omes
to
biom
e
tr
ic
identi
f
ic
a
ti
on,
3D
im
a
ging
r
e
pr
e
s
e
nts
a
s
igni
f
ica
nt
a
dva
nc
e
ment
o
ve
r
tr
a
dit
ional
2D
methods
,
s
e
tt
ing
a
ne
w
s
tan
da
r
d
f
o
r
a
c
c
ur
a
c
y,
s
e
c
ur
it
y,
a
nd
c
onve
nienc
e
in
the
b
iom
e
t
r
ics
f
ield.
AC
KNOWL
E
DGE
M
E
NT
We
would
li
ke
to
e
xpr
e
s
s
our
he
a
r
tf
e
lt
thanks
to
o
ur
guide
f
or
he
r
unwa
ve
r
ing
guidanc
e
,
invalua
ble
ins
ight
s
,
a
nd
e
nc
our
a
ge
ment
thr
oughout
the
r
e
s
e
a
r
c
h
pr
oc
e
s
s
.
W
e
would
a
ls
o
li
ke
to
thank
Nitt
e
M
e
e
na
ks
hi
I
ns
ti
tut
e
of
T
e
c
hnology
f
or
p
r
ovidi
ng
the
ne
c
e
s
s
a
r
y
s
uppor
t
f
o
r
c
a
r
r
ying
out
thi
s
r
e
s
e
a
r
c
h.
F
UN
DI
NG
I
NF
ORM
AT
I
ON
T
he
a
uthor
s
would
li
ke
to
in
f
or
m
that
thi
s
p
r
ojec
t
ha
s
be
e
n
f
unde
d
by
the
Ka
r
na
taka
S
tate
C
ounc
il
f
or
S
c
ienc
e
a
nd
T
e
c
hnology
(
KSC
S
T
)
B
a
nga
lor
e
.
AU
T
HO
R
CONT
RI
B
U
T
I
ONS
S
T
AT
E
M
E
N
T
T
his
jour
na
l
us
e
s
the
C
ontr
ibut
o
r
R
oles
T
a
xo
nomy
(
C
R
e
diT
)
to
r
e
c
ognize
indi
vidual
a
uthor
c
ontr
ibut
ions
,
r
e
duc
e
a
utho
r
s
hip
dis
putes
,
a
nd
f
a
c
il
it
a
te
c
oll
a
bor
a
ti
on.
T
he
a
utho
r
s
would
li
ke
to
in
f
or
m
that
the
tea
m
c
ons
is
ted
of
f
ou
r
membe
r
s
,
e
a
c
h
of
who
m
c
ontr
ibu
ted
mea
ningf
u
ll
y
to
t
he
ove
r
a
ll
de
ve
lopm
e
nt
of
the
pa
pe
r
.
T
he
de
tailed
r
ubr
ics
us
e
d
f
o
r
the
a
na
lys
is
a
r
e
e
nc
los
e
d
he
r
e
in.
Nam
e
of
Au
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Divya
Ga
nga
c
ha
nna
iah
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
M
a
matha
Ar
uva
na
ll
i
S
hivar
a
j
✓
✓
✓
✓
✓
✓
✓
✓
✓
Honga
nur
C
ha
ndr
a
s
e
kha
r
a
iah
Na
ga
r
a
j
✓
✓
✓
✓
✓
✓
✓
✓
✓
P
r
a
s
a
nna
Gur
ur
a
j
P
a
ga
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
li
z
a
ti
on
M
:
M
e
th
odol
ogy
So
:
So
f
twa
r
e
Va
:
Va
li
da
ti
on
Fo
:
Fo
r
ma
l
a
na
ly
s
is
I
:
I
nve
s
ti
ga
ti
on
R
:
R
e
s
our
c
e
s
D
:
D
a
ta
C
ur
a
ti
on
O
:
W
r
it
in
g
-
O
r
ig
in
a
l
D
r
a
f
t
E
:
W
r
it
in
g
-
R
e
vi
e
w
&
E
di
ti
ng
Vi
:
Vi
s
ua
li
z
a
ti
on
Su
:
Su
pe
r
vi
s
io
n
P
:
P
r
oj
e
c
t
a
dmi
ni
s
tr
a
ti
on
Fu
:
Fu
ndi
ng a
c
qui
s
it
io
n
CONF
L
I
CT
OF
I
NT
E
RE
S
T
S
T
AT
E
M
E
N
T
T
he
a
uthor
s
would
l
ike
to
de
c
lar
e
that
ther
e
is
no
c
onf
li
c
t
of
int
e
r
e
s
t
pe
r
taining
to
f
inanc
ial,
pe
r
s
ona
l,
or
pr
o
f
e
s
s
ional
in
c
onne
c
ti
on
with
the
manus
c
r
ipt
s
.
I
NF
ORM
E
D
CONSE
NT
T
he
a
uthor
s
would
li
ke
to
inf
or
m
that
the
da
ta
s
e
ts
ha
ve
be
e
n
take
n
with
pr
io
r
c
ons
e
nt
f
r
om
Hong
Kong
P
olyt
e
c
hnic
univer
s
it
y.
DA
T
A
AV
AI
L
A
B
I
L
I
T
Y
T
he
a
uthor
s
would
li
ke
to
inf
or
m
that
the
da
ta
s
e
t
f
or
c
a
r
r
ying
out
the
r
e
s
e
a
r
c
h
is
publi
c
ly
a
va
il
a
ble
in
htt
ps
:/
/www
4.
c
omp.
polyu
.
e
du.
hk/~c
s
a
jaykr
/3D
Knuc
kle.
htm
.
O
f
f
icia
l
c
ons
e
nt
f
o
r
us
ing
the
da
taba
s
e
wa
s
obtaine
d
f
r
om
the
Hong
Kong
P
olyt
e
c
hnic
unive
r
s
it
y.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
Ar
ti
f
I
ntell
I
S
S
N:
2252
-
8938
Sur
v
e
y
on
3D
biome
tr
ic
tr
ait
s
for
human
identif
ica
ti
on
(
Div
y
a
Gangac
hannaiah)
3151
RE
F
E
RE
NC
E
S
[
1]
M
.
H
a
mm
a
d,
Y
.
L
iu
,
a
nd
K
.
W
a
ng,
“
M
ul
ti
moda
l
bi
ome
tr
ic
a
ut
he
nt
ic
a
ti
on
s
ys
te
ms
us
in
g
c
onvolut
io
n
ne
ur
a
l
ne
twor
k
ba
s
e
d
on
di
f
f
e
r
e
nt
le
ve
l
f
us
io
n
o
f
E
C
G
a
nd
f
in
ge
r
pr
in
t,
”
I
E
E
E
A
c
c
e
s
s
,
vol
.
7,
pp.
25527
–
25542,
2019,
doi
:
10.1109/AC
C
E
S
S
.2018.2886573.
[
2]
R
.
D
a
s
,
E
.
P
ic
iu
c
c
o,
E
.
M
a
io
r
a
na
,
a
nd
P
.
C
a
mpi
s
i,
“
C
onvolut
io
na
l
ne
ur
a
l
ne
twor
k
f
or
f
in
ge
r
-
ve
in
-
ba
s
e
d
bi
ome
tr
ic
id
e
nt
if
ic
a
ti
on,”
I
E
E
E
T
r
ans
ac
ti
ons
on
I
nf
or
m
at
io
n
F
or
e
ns
ic
s
and
Se
c
ur
it
y
,
vol
.
14,
no.
2,
pp.
360
–
373,
F
e
b.
2018,
doi
:
10.1109/T
I
F
S
.2018.2850320.
[
3]
A
.
T
a
r
a
nnum,
Z
.
U
.
R
a
hma
n,
L
.
K
.
R
a
o,
T
.
S
r
in
iv
a
s
ul
u,
a
nd
A
.
L
a
y
-
E
kua
ki
ll
e
,
“
A
n
e
f
f
ic
ie
nt
mul
ti
-
moda
l
bi
ome
tr
ic
s
e
ns
in
g
a
nd
a
ut
he
nt
ic
a
ti
on
f
r
a
me
w
or
k
f
or
di
s
tr
ib
ut
e
d
a
ppl
ic
a
ti
ons
,”
I
E
E
E
Se
ns
or
s
J
our
nal
,
vol
.
20,
no.
24,
pp.
15014
–
15025,
2020,
doi
:
10.1109/J
S
E
N
.2020.3012536.
[
4]
G
.
Y
.
I
z
a
de
e
n
a
nd
S
.
Y
.
A
me
e
n,
“
S
ma
r
t
a
ndr
oi
d
gr
a
phi
c
a
l
p
a
s
s
w
or
d
s
tr
a
te
gy:
a
r
e
vi
e
w
,”
A
s
ia
n
J
ou
r
nal
of
R
e
s
e
ar
c
h
in
C
om
p
ut
e
r
Sc
ie
nc
e
, pp. 59
–
69, 2021, doi:
10.9734/ajr
c
os
/2
021/
v9i
230220.
[
5]
P
.
P
uj
a
r
,
A
.
K
uma
r
,
a
nd
V
.
K
uma
r
,
“
P
la
nt
le
a
f
de
te
c
ti
on
th
r
o
ugh
ma
c
hi
ne
le
a
r
ni
ng
ba
s
e
d
im
a
ge
c
la
s
s
if
ic
a
ti
on
a
ppr
oa
c
h,”
I
A
E
S
I
nt
e
r
nat
io
nal
J
our
nal
of
A
r
ti
fi
c
ia
l
I
nt
e
ll
ig
e
nc
e
(IJ
-
A
I
)
,
vol
.
13,
no.
1,
pp.
1139
–
1148,
M
a
r
.
2024,
doi
:
10.11591/i
ja
i.
v13.i1.pp1139
-
1148.
[
6]
S
.
H
.
S
r
e
e
dha
r
a
,
V
.
K
uma
r
,
a
nd
S
.
S
a
lm
a
,
“
E
f
f
ic
ie
nt
bi
g
da
t
a
c
lu
s
te
r
in
g
us
in
g
a
dhoc
f
uz
z
y
c
me
a
ns
a
nd
a
ut
o
-
e
nc
od
e
r
C
N
N
,”
L
e
c
tu
r
e
N
ot
e
s
i
n N
e
tw
o
r
k
s
and S
y
s
te
m
s
, vol
. 563, pp. 353
–
368,
2023, doi:
10.1007/978
-
981
-
19
-
7402
-
1_25.
[
7]
Q
.
Z
ha
ng,
“
D
e
e
p
l
e
a
r
ni
ng
of
e
le
c
tr
oc
a
r
di
ogr
a
phy
dyna
mi
c
s
f
or
b
io
me
tr
ic
huma
n
id
e
nt
if
ic
a
ti
on
in
e
r
a
of
I
oT
,”
in
2018
9t
h
I
E
E
E
A
nnual
U
bi
qui
to
us
C
om
put
in
g,
E
le
c
tr
oni
c
s
and
M
obi
le
C
om
m
uni
c
at
io
n
C
onf
e
r
e
n
c
e
,
U
E
M
C
O
N
2018
,
2018,
pp.
885
–
888,
doi
:
10.1109/UE
M
C
O
N
.2018.8796676.
[
8]
K
.
H
.
M
.
C
he
ng
a
nd
A
.
K
uma
r
,
“
C
ont
a
c
tl
e
s
s
bi
o
me
tr
ic
id
e
nt
if
ic
a
ti
on
us
in
g
3D
f
in
ge
r
knuc
kl
e
pa
tt
e
r
ns
,”
I
E
E
E
T
r
ans
ac
ti
ons
on
P
at
te
r
n A
nal
y
s
is
and M
ac
hi
ne
I
nt
e
ll
ig
e
nc
e
, vol
. 42, no. 8, pp. 1
868
–
1883, 2020, doi:
10.1109/T
P
A
M
I
.2019.2904232.
[
9]
M
.
B
a
s
s
i
a
nd
P
.
T
r
iv
e
r
bi
,
“
H
uma
n
bi
ome
tr
ic
id
e
nt
if
ic
a
ti
on
th
r
ough
br
a
in
pr
in
t,
”
in
P
r
oc
e
e
di
ngs
of
th
e
2nd
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
E
le
c
tr
oni
c
s
,
C
om
m
uni
c
at
io
n
and
A
e
r
o
s
p
ac
e
T
e
c
hnol
ogy
,
I
C
E
C
A
2018
,
2018,
pp.
1514
–
1518,
doi
:
10.1109/I
C
E
C
A
.2018.8474646.
[
10]
R
.
S
.
K
uz
u,
E
.
P
ic
iu
c
c
o,
E
.
M
a
io
r
a
na
,
a
nd
P
.
C
a
mpi
s
i,
“
O
n
-
t
he
-
f
ly
f
in
ge
r
-
ve
in
-
ba
s
e
d
bi
ome
tr
ic
r
e
c
ogni
ti
on
us
in
g
de
e
p
ne
ur
a
l
ne
twor
ks
,”
I
E
E
E
T
r
ans
ac
ti
ons
on
I
nf
or
m
at
io
n
F
or
e
n
s
ic
s
and
S
e
c
ur
it
y
,
vol
.
15,
pp.
2641
–
2654,
2020,
doi
:
10.1109/T
I
F
S
.2020.2971144.
[
11]
G
. J
a
s
w
a
l,
A
. K
a
ul
, a
nd R
. N
a
th
, “
M
ul
ti
moda
l
bi
ome
tr
ic
a
ut
he
n
ti
c
a
ti
on s
ys
te
m us
in
g ha
nd s
ha
p
e
, pa
lm
pr
in
t,
a
nd ha
nd ge
ome
tr
y,”
A
dv
anc
e
s
i
n I
nt
e
ll
ig
e
nt
Sy
s
te
m
s
and C
om
put
in
g
, vol
. 799, pp. 5
57
–
570, 2019, doi:
10.1007/978
-
981
-
13
-
1135
-
2_42.
[
12]
K
.
P
r
ih
odova
a
nd
M
.
H
ub,
“
B
io
me
tr
ic
pr
iv
a
c
y
th
r
ough
ha
nd
g
e
ome
tr
y
-
a
s
ur
ve
y,”
in
P
r
oc
e
e
di
ngs
of
th
e
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on I
nf
or
m
at
io
n and Digit
al
T
e
c
hnol
ogi
e
s
2019, I
D
T
2019
, 2019, pp. 395
–
401, doi:
10.1109/DT
.2019.8813660.
[
13]
N
.
W
a
ng
e
t
al
.
,
“
3D
r
e
c
on
s
tr
uc
ti
on
a
nd
s
e
gme
nt
a
ti
on
s
ys
te
m
f
or
pa
ve
me
nt
pot
hol
e
s
ba
s
e
d
on
im
pr
ove
d
s
tr
uc
tu
r
e
-
f
r
om
-
mot
io
n
(
S
F
M
)
a
nd de
e
p l
e
a
r
ni
ng,”
C
ons
tr
u
c
ti
on and B
ui
ld
in
g M
at
e
r
ia
ls
, vol
. 398, 2023, doi:
10.1016/j
.c
onbuil
dma
t.
2023.132499.
[
14]
W
.
K
a
ng,
H
.
L
iu
,
W
.
L
uo,
a
nd
F
.
D
e
ng,
“
S
tu
dy
of
a
f
ul
l
-
vi
e
w
3D
f
in
ge
r
ve
in
ve
r
if
ic
a
ti
on
te
c
hni
que
,”
I
E
E
E
T
r
ans
ac
ti
ons
on
I
nf
or
m
at
io
n F
or
e
ns
ic
s
and Se
c
ur
it
y
, vol
. 15, pp. 1175
–
1189, 20
20, doi:
10.1109/T
I
F
S
.2019.2928507.
[
15]
D
.
H
.
A
l
-
J
a
na
bi
a
nd
A
.
M
.
A
l
-
J
uboor
i,
“
3D
-
f
in
ge
r
knuc
kl
e
r
e
c
ogni
t
io
n
us
in
g
c
onvolut
io
na
l
ne
ur
a
l
ne
twor
k,”
in
P
r
oc
e
e
di
n
gs
-
C
SC
T
I
T
2022:
5t
h
C
ol
le
ge
of
Sc
ie
nc
e
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
R
e
c
e
nt
T
r
e
nd
s
in
I
nf
or
m
at
io
n
T
e
c
hnol
og
y
,
2022,
pp.
175
–
178,
doi
:
10.1109/C
S
C
T
I
T
56299.2022.10145675.
[
16]
S
.
C
he
n,
Z
.
G
uo,
J
.
F
e
ng,
a
nd
J
.
Z
hou,
“
A
n
im
pr
ove
d
c
ont
a
c
t
-
ba
s
e
d
hi
gh
-
r
e
s
ol
ut
io
n
pa
lm
pr
in
t
im
a
ge
a
c
qui
s
it
io
n
s
ys
te
m,”
I
E
E
E
T
r
ans
ac
ti
ons
on I
ns
t
r
um
e
nt
at
io
n and M
e
as
ur
e
m
e
nt
, vol
. 69, no.
9, pp. 6816
–
6827, 2020, doi:
10.1109/T
I
M
.2020.2976081.
[
17]
K
. H
. M
. C
he
ng a
nd A
. K
uma
r
, “
E
f
f
ic
ie
nt
a
nd
a
c
c
ur
a
te
3D
f
in
g
e
r
knuc
kl
e
ma
tc
hi
ng us
in
g s
ur
f
a
c
e
ke
y point
s
,”
I
E
E
E
T
r
ans
ac
ti
ons
on I
m
age
P
r
oc
e
s
s
in
g
, vol
. 29, pp. 8903
–
8915, 2020, doi:
10.11
09/
T
I
P
.2020.3021294.
[
18]
A
. K
uma
r
, “
T
he
H
ong Kong Pol
yt
e
c
hni
c
U
ni
ve
r
s
it
y
c
ont
a
c
tl
e
s
s
3D
f
in
ge
r
K
nuc
kl
e
I
ma
ge
s
da
ta
ba
s
e
,
”
T
he
H
ong K
ong P
ol
y
te
c
hni
c
U
ni
v
e
r
s
it
y
, 2019.
[
O
nl
in
e
]
. A
va
il
a
bl
e
:
ht
tp
s
:/
/ww
w
4.
c
omp.pol
yu.e
du.hk/~
c
s
a
j
a
ykr
/3
D
K
nuc
kl
e
.ht
m
[
19]
K
.
H
.
M
.
C
he
ng
a
nd
A
.
K
um
a
r
,
“
A
c
c
ur
a
te
3D
f
in
ge
r
knuc
kl
e
r
e
c
ogni
ti
on
us
in
g
a
ut
o
-
ge
ne
r
a
te
d
s
im
il
a
r
it
y
f
unc
ti
ons
,”
I
E
E
E
T
r
ans
ac
ti
ons
on
B
io
m
e
tr
ic
s
,
B
e
hav
io
r
,
and
I
de
nt
it
y
Sc
ie
nc
e
,
vol
.
3,
no.
2,
pp.
203
–
213,
2021,
doi
:
10.1109/T
B
I
O
M
.2021.3051062.
[
20]
D
.
P
a
l
ma
,
P
.
L
.
M
ont
e
s
s
or
o,
G
.
G
io
r
da
no,
a
nd
F
.
B
la
nc
hi
ni
,
“
B
io
me
tr
ic
pa
lm
pr
in
t
ve
r
if
ic
a
ti
on:
a
dyna
mi
c
a
l
s
ys
t
e
m
a
ppr
oa
c
h,”
I
E
E
E
T
r
ans
ac
ti
ons
on
Sy
s
te
m
s
,
M
an,
and
C
y
be
r
ne
ti
c
s
:
Sy
s
te
m
s
,
vol
.
49,
no.
12,
pp.
2676
–
2687,
2019,
doi
:
10.1109/T
S
M
C
.2017.2771232.
[
21]
G.
J
a
s
w
a
l
a
nd
R
.
C
.
P
ooni
a
,
“
S
e
le
c
ti
on
of
opt
im
iz
e
d
f
e
a
tu
r
e
s
f
or
f
us
io
n
of
pa
lm
pr
in
t
a
nd
f
in
ge
r
knuc
kl
e
-
ba
s
e
d
pe
r
s
on
a
ut
he
nt
ic
a
ti
on,”
E
x
pe
r
t
Sy
s
te
m
s
, vol
. 38, no. 1, 2021, doi
:
10.11
11/
e
xs
y.12523.
[
22]
K
.
H
.
M
.
C
he
ng
a
nd
A
.
K
uma
r
, “
D
e
e
p
f
e
a
tu
r
e
c
ol
la
bor
a
ti
o
n
f
or
c
ha
ll
e
ngi
ng
3D
f
in
ge
r
knuc
kl
e
id
e
nt
if
ic
a
ti
on,”
I
E
E
E
T
r
ans
ac
ti
ons
on I
nf
or
m
at
io
n F
or
e
ns
ic
s
and Se
c
ur
it
y
, vol
. 16, pp. 1158
–
1173,
2021, doi:
10.1109/T
I
F
S
.2020.3029906.
[
23]
S
. A
r
or
a
, M
. P
.
S
. B
ha
ti
a
, a
nd H
. K
ukr
e
ja
, “
A
mul
ti
moda
l
bi
om
e
tr
ic
s
ys
te
m f
or
s
e
c
ur
e
us
e
r
i
de
nt
if
ic
a
ti
on ba
s
e
d on de
e
p l
e
a
r
ni
ng,”
P
r
oc
e
e
di
ngs
of
F
if
th
I
nt
e
r
nat
io
nal
C
ongr
e
s
s
on
I
nf
or
m
at
io
n
and
C
om
m
uni
c
at
io
n
T
e
c
hnol
ogy
(
I
C
I
C
T
20
20)
,
O
c
t.
2020,
pp. 95
–
103, doi:
10.1007/978
-
981
-
15
-
5856
-
6_8.
[
24]
M
.
O
.
O
lo
ye
de
a
nd
G
.
P
.
H
a
nc
ke
,
“
U
ni
moda
l
a
nd
mul
ti
mo
da
l
bi
ome
tr
ic
s
e
ns
in
g
s
ys
te
m
s
:
a
r
e
vi
e
w
,”
I
E
E
E
A
c
c
e
s
s
,
vol
.
4,
pp. 7532
–
7555, 2016, doi:
10.1109/AC
C
E
S
S
.2016.2614720.
[
25]
B
.
Y
a
ng,
X
.
X
ia
ng,
D
.
X
u,
X
.
W
a
ng,
a
nd
X
.
Y
a
ng,
“
3D
pa
lm
p
r
in
t
r
e
c
ogni
ti
on
us
in
g
s
ha
pe
in
de
x
r
e
pr
e
s
e
nt
a
ti
on
a
nd
f
r
a
gi
le
bi
ts
,”
M
ul
ti
m
e
di
a T
ool
s
and A
ppl
ic
at
io
ns
, vol
. 76, no. 14, pp. 15357
–
15375, 2017, doi:
10.1007/s
11042
-
016
-
3832
-
1.
[
26]
R
.
S
.
K
uz
u,
E
.
M
a
io
r
a
na
,
a
nd
P
.
C
a
mpi
s
i,
“
G
e
nde
r
-
s
pe
c
if
ic
c
ha
r
a
c
te
r
is
ti
c
s
f
or
ha
nd
-
ve
in
bi
ome
tr
ic
r
e
c
ogni
ti
on:
a
n
a
ly
s
is
a
nd
e
xpl
oi
ta
ti
on,
”
I
E
E
E
A
c
c
e
s
s
, vol
. 11, pp. 11700
–
11710, 2023, d
oi
:
10.1109/AC
C
E
S
S
.2023.3239894.
[
27]
G
.
D
iv
ya
,
G
.
P
.
P
r
a
s
a
nna
,
M
.
M
a
dhus
hr
e
e
,
A
.
A
c
hut
,
a
nd
V
is
hw
a
,
“
A
d
a
pt
iv
e
li
ne
a
r
in
te
r
pol
a
ti
on
a
lg
or
it
hm
f
or
2D
D
I
C
O
M
mul
ti
pl
e
s
kul
l
im
a
ge
s
,”
in
2024
5t
h
I
E
E
E
G
lo
bal
C
onf
e
r
e
nc
e
f
or
A
dv
anc
e
m
e
nt
in
T
e
c
hnol
og
y
(
G
C
A
T
)
,
O
c
t.
2024,
vol
.
11,
no.
1,
pp. 1
–
5, doi:
10.1109/G
C
A
T
62922.2024.10924040.
[
28]
M
.
S
.
L
ohi
th
,
Y
.
S
.
K
.
M
a
nj
una
th
,
a
nd
M
.
N
.
E
s
hw
a
r
a
ppa
,
“
M
ul
ti
moda
l
bi
ome
tr
ic
pe
r
s
on
a
ut
he
nt
ic
a
ti
on
us
in
g
f
a
c
e
,
e
a
r
a
nd
pe
r
io
c
ul
a
r
r
e
gi
on
ba
s
e
d
on
c
onvolut
io
n
ne
ur
a
l
ne
twor
ks
,”
I
nt
e
r
nat
io
nal
J
our
nal
of
I
m
age
and
G
r
aphi
c
s
,
vol
.
23,
no.
2,
2023,
doi
:
10.1142/S
0219467823500195.
[
29]
Ö
.
B
in
göl
a
nd
M
.
E
ki
nc
i,
“
S
te
r
e
o
-
ba
s
e
d
pa
lm
pr
in
t
r
e
c
ogni
ti
on
in
va
r
io
us
3D
pos
tu
r
e
s
,”
E
x
pe
r
t
Sy
s
te
m
s
w
it
h
A
p
pl
ic
at
io
ns
,
vol
. 78, pp. 74
–
88, 2017, doi:
10.1016/j
.e
s
w
a
.2017.01.025.
[
30]
E
.
A
l
A
lk
e
e
m
e
t
al
.
,
“
R
obu
s
t
de
e
p
id
e
nt
if
ic
a
ti
on
us
in
g
E
C
G
a
n
d
mul
ti
moda
l
bi
ome
tr
ic
s
f
or
in
dus
tr
ia
l
in
te
r
ne
t
of
th
in
gs
,”
A
d
H
oc
N
e
tw
or
k
s
, vol
. 121, 2021, doi:
10.1016/j
.a
dhoc
.2021
.102581.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
4
,
Augus
t
2025
:
314
3
-
3152
3152
[
31]
S
.
L
i,
B
.
Z
ha
ng,
L
.
F
e
i,
S
.
Z
ha
o,
a
nd
Y
.
Z
hou,
“
L
e
a
r
ni
ng
s
pa
r
s
e
a
nd
di
s
c
r
im
in
a
ti
ve
mul
ti
moda
l
f
e
a
tu
r
e
c
ode
s
f
or
f
in
ge
r
r
e
c
ogni
ti
on,”
I
E
E
E
T
r
ans
ac
ti
ons
on M
ul
ti
m
e
di
a
, vol
. 25, pp. 805
–
815, 2023, doi:
10.1109/T
M
M
.2021.3132166.
[
32]
S
.
L
i
a
nd
B
.
Z
ha
ng,
“
J
oi
nt
di
s
c
r
im
in
a
ti
ve
s
pa
r
s
e
c
odi
ng
f
or
r
obus
t
ha
nd
-
ba
s
e
d
mul
ti
moda
l
r
e
c
ogni
ti
on,”
I
E
E
E
T
r
ans
ac
ti
ons
on
I
nf
or
m
at
io
n F
or
e
ns
ic
s
and Se
c
ur
it
y
, vol
. 16, pp. 3186
–
3198, 20
21, doi:
10.1109/T
I
F
S
.2021.3074315.
[
33]
W
.
Y
a
ng,
Z
.
C
he
n,
J
.
H
ua
ng,
L
.
W
a
ng,
a
nd
W
.
K
a
ng,
“
L
F
M
B
-
3D
F
B
:
a
la
r
ge
-
s
c
a
l
e
f
in
ge
r
mul
ti
-
bi
ome
tr
ic
da
ta
ba
s
e
a
nd
be
nc
hma
r
k f
or
3D
f
in
ge
r
bi
ome
tr
ic
s
,”
2021, doi:
10.1109/I
J
C
B
52358.2021.9484369.
[
34]
W
.
Y
a
ng,
Z
.
C
he
n,
J
.
H
ua
ng,
a
nd
W
.
K
a
ng,
“
A
nov
e
l
s
y
s
te
m
a
nd
e
xp
e
r
im
e
nt
a
l
s
tu
dy
f
or
3D
f
in
ge
r
mul
ti
b
io
me
tr
ic
s
,”
I
E
E
E
T
r
ans
ac
ti
ons
on
B
io
m
e
tr
ic
s
,
B
e
hav
io
r
,
and
I
de
nt
it
y
Sc
ie
nc
e
,
vol
.
4,
no.
4,
pp.
471
–
485,
2022,
doi
:
10.1109/T
B
I
O
M
.2022.3181121.
B
I
OG
RA
P
HI
E
S
OF
AU
T
HO
RS
D
i
v
y
a
G
a
ng
a
cha
nn
a
i
a
h
i
s
cu
rren
t
l
y
w
o
r
k
i
n
g
as
a
n
A
s
s
i
s
t
a
n
t
Pr
o
fes
s
o
r
G
rad
III
i
n
E
l
ec
t
ro
n
i
c
s
an
d
Co
mm
u
n
i
cat
i
o
n
E
n
g
i
n
eeri
n
g
o
f
N
i
t
t
e
Meen
ak
s
h
i
In
s
t
i
t
u
t
e
o
f
T
ech
n
o
l
o
g
y
,
Y
el
a
h
an
k
a,
an
d
B
an
g
al
o
re.
Sh
e
co
m
p
l
e
t
ed
h
er
B
.
E
.
i
n
E
l
ec
t
ro
n
i
c
s
an
d
Co
mm
u
n
i
cat
i
o
n
E
n
g
i
n
eeri
n
g
fro
m
E
a
s
t
Po
i
n
t
Co
l
l
eg
e
o
f
E
n
g
i
n
eer
i
n
g
an
d
T
ech
n
o
l
o
g
y
Ban
g
al
o
re
an
d
M
.
T
ech
.
i
n
D
i
g
i
t
a
l
E
l
ect
r
o
n
i
cs
fr
o
m
E
a
s
t
W
e
s
t
I
n
s
t
i
t
u
t
e
o
f
T
ech
n
o
l
o
g
y
Ban
g
al
o
re.
Sh
e
h
a
s
1
7
y
ear
s
o
f
t
each
i
n
g
ex
p
eri
e
n
ce.
Pres
e
n
t
l
y
p
u
r
s
i
n
g
h
e
r
Ph
.
D
.
i
n
N
i
t
t
e
U
n
i
v
ers
i
t
y
Man
g
al
o
re.
H
er
area
s
o
f
i
n
t
eres
t
are
s
i
g
n
a
l
p
ro
ce
s
s
i
n
g
an
d
i
mag
e
p
ro
ce
s
s
i
n
g
.
Sh
e
can
b
e
co
n
t
ac
t
ed
a
t
emai
l
:
d
i
v
y
a
g
_
1
2
@
red
i
ffmai
l
.
co
m.
M
a
m
a
th
a
A
ruv
a
n
a
l
l
i
Shi
v
a
r
a
j
i
s
cu
rren
t
l
y
w
o
rk
i
n
g
as
A
s
s
o
ci
a
t
e
Pro
fes
s
o
r
i
n
t
h
e
D
ep
ar
t
men
t
o
f
E
l
ect
r
o
n
i
cs
an
d
Co
mmu
n
i
ca
t
i
o
n
E
n
g
i
n
eer
i
n
g
at
N
MA
M
In
s
t
i
t
u
t
e
o
f
T
ech
n
o
l
o
g
y
,
N
i
t
t
e,
U
d
u
p
i
,
K
arn
a
t
ak
a,
I
n
d
i
a.
Sh
e
h
a
s
2
5
y
ear
s
o
f
t
each
i
n
g
ex
p
er
i
en
ce.
Sh
e
i
s
t
h
e
au
t
h
o
r
o
f
n
i
n
e
i
n
t
ern
a
t
i
o
n
a
l
j
o
u
rn
a
l
s
an
d
s
i
x
i
n
t
e
rn
at
i
o
n
al
co
n
feren
ce
s
i
n
t
h
e
fi
el
d
o
f
mu
l
t
i
s
p
ec
t
ral
i
mag
e
co
m
p
res
s
i
o
n
.
Sh
e
i
s
t
h
e
au
t
h
o
r
o
f
t
h
e
‘
N
et
w
o
rk
t
h
eo
r
y
’
,
‘
E
n
g
i
n
eer
i
n
g
s
t
a
t
i
s
t
i
cs
an
d
l
i
n
ear
a
l
g
e
b
ra’,
an
d
‘
C
o
n
t
ro
l
e
n
g
i
n
eeri
n
g
’
.
H
er
area
s
o
f
i
n
t
ere
s
t
are
s
i
g
n
a
l
p
ro
ce
s
s
i
n
g
,
i
mag
e
c
o
mp
re
s
s
i
o
n
,
an
d
c
o
n
t
ro
l
en
g
i
n
eeri
n
g
.
Sh
e
i
s
a
s
e
n
i
o
r
memb
er
o
f
IE
E
E
.
Sh
e
can
b
e
co
n
t
ac
t
ed
a
t
emai
l
:
mamat
h
ag
_
1
2
@
red
i
ffma
i
l
.
co
m.
Ho
ng
a
nur
C
h
a
ndra
s
ekha
r
a
i
a
h
N
a
g
a
r
a
j
i
s
cu
rren
t
l
y
w
o
rk
i
n
g
as
t
h
e
Pri
n
ci
p
al
o
f
N
MIT
,
Ban
g
al
o
re.
H
e
h
o
l
d
s
B.
E
.
(E
l
ect
r
o
n
i
c
s
an
d
Co
mmu
n
i
ca
t
i
o
n
s
)
d
e
g
ree
fro
m
t
h
e
U
n
i
v
er
s
i
t
y
o
f
My
s
o
re,
M.
E
.
(Co
mm
u
n
i
cat
i
o
n
Sy
s
t
e
ms
)
d
eg
ree
fr
o
m
P.
S.
G
Co
l
l
eg
e
o
f
T
ech
n
o
l
o
g
y
,
Co
i
mb
a
t
o
re
an
d
Ph
.
D
.
(Bi
o
med
i
ca
l
Si
g
n
al
Pr
o
ces
s
i
n
g
an
d
In
s
t
ru
me
n
t
a
t
i
o
n
)
fr
o
m
In
d
i
an
In
s
t
i
t
u
t
e
o
f
T
ech
n
o
l
o
g
y
,
Ch
e
n
n
a
i
.
H
e
h
as
t
each
i
n
g
ex
p
eri
e
n
ce
o
f
al
m
o
s
t
4
d
eca
d
es
,
h
i
s
ai
ms
t
o
b
r
i
n
g
N
MI
T
amo
n
g
t
h
e
t
o
p
5
0
ed
u
cat
i
o
n
i
n
s
t
i
t
u
t
es
acc
o
rd
i
n
g
t
o
N
IRF
ran
k
i
n
g
.
H
e
h
a
s
mas
s
i
v
e
ex
p
eri
e
n
ce
o
f
4
2
y
ear
s
i
n
t
ea
c
h
i
n
g
.
H
e
h
as
v
i
s
i
t
e
d
1
5
co
u
n
t
r
i
es
a
n
d
s
t
u
d
i
ed
t
h
e
U
n
i
v
er
s
i
t
y
E
n
g
i
n
eeri
n
g
E
d
u
cat
i
o
n
Sy
s
t
em,
i
n
v
o
l
v
i
n
g
v
ari
o
u
s
l
a
b
o
ra
t
o
r
i
es
,
a
n
d
ce
n
t
er
s
o
f
ex
cel
l
en
ce
.
H
e
h
a
s
p
u
b
l
i
s
h
ed
8
0
p
a
p
ers
n
a
t
i
o
n
a
l
l
y
an
d
i
n
t
ern
a
t
i
o
n
a
l
l
y
.
H
e
can
b
e
co
n
t
act
e
d
at
emai
l
:
p
ri
n
ci
p
al
@
n
m
i
t
.
ac.
i
n
.
P
ra
s
a
nna
Gurur
a
j
P
a
g
a
i
s
cu
rre
n
t
l
y
w
o
r
k
i
n
g
a
s
an
A
s
s
o
ci
a
t
e
Pro
fe
s
s
o
r
i
n
D
ep
ar
t
men
t
o
f
E
l
ect
r
o
n
i
cs
an
d
Co
mmu
n
i
cat
i
o
n
E
n
g
i
n
eeri
n
g
at
N
i
t
t
e
Meen
ak
s
h
i
In
s
t
i
t
u
t
e
o
f
T
ech
n
o
l
o
g
y
,
Y
el
a
h
an
k
a,
Ban
g
al
o
re.
H
e
c
o
mp
l
et
e
d
h
i
s
B
.
E
.
i
n
E
l
ec
t
ro
n
i
c
s
a
n
d
C
o
mmu
n
i
ca
t
i
o
n
E
n
g
i
n
eeri
n
g
fr
o
m
P
D
A
C
o
l
l
eg
e
o
f
E
n
g
i
n
eeri
n
g
G
u
l
b
arg
a
an
d
M
.
T
ech
.
i
n
In
d
u
s
t
r
i
al
E
l
ec
t
ro
n
i
c
s
fro
m
SJ
CE
M
y
s
o
re
an
d
P
h
.
D
.
fro
m
U
n
i
v
er
s
i
t
y
o
f
My
s
o
r
e
o
n
A
n
t
en
n
as
.
H
e
h
as
o
v
er
2
0
p
u
b
l
i
cat
i
o
n
s
i
n
rep
u
t
e
d
j
o
u
rn
a
l
s
an
d
c
o
n
feren
ce
s
.
H
e
h
as
o
v
er
1
6
y
ear
s
o
f
t
each
i
n
g
ex
p
eri
e
n
ce
.
H
e
can
b
e
co
n
t
ac
t
ed
a
t
emai
l
:
p
ra
s
an
n
a.
p
a
g
a@
n
mi
t
.
ac.
i
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.