T
E
L
KO
M
NIK
A
, V
ol
.
17
,
No.
5,
O
c
tob
er
20
1
9,
p
p.
2
55
4
~
25
60
IS
S
N: 1
69
3
-
6
93
0
,
accr
ed
ited
F
irst
Gr
ad
e b
y K
em
en
r
istekdikti,
Decr
ee
No: 2
1/E/
K
P
T
/20
18
DOI:
10.12928/TE
LK
OM
N
IK
A
.v
1
7
i
5
.
12799
◼
25
54
Rec
ei
v
ed
O
c
tob
er
5
, 2
01
8
;
Rev
i
s
ed
Ma
r
c
h
28
, 2
01
9
; A
c
c
ep
ted
A
pri
l
12
, 2
01
9
A
n
o
v
e
l ele
men
t
ar
y
s
pat
i
al ex
pa
nd
in
g sc
hem
e f
orm
o
n
SIS
R
met
ho
d
w
i
th
m
od
i
f
y
in
g
G
e
man
&
McCl
ure
f
un
ct
ion
Dar
u
n
Ke
sr
ar
at
*
1
, Ko
r
n
ka
mo
l T
h
ak
u
lsukanant
2
,
V
o
r
apo
j P
at
anav
ijit
3
As
s
u
m
p
t
i
o
n
Un
i
v
e
rs
i
ty
o
f
T
h
a
i
l
a
n
d
(AU),
T
h
a
i
l
a
n
d
*C
o
rre
s
p
o
n
d
i
n
g
a
u
th
o
r,
e
-
m
a
i
l
:
d
a
ru
n
k
s
r@g
m
a
i
l
.
c
o
m
1
,
k
th
a
k
u
l
s
u
k
a
n
a
n
t@y
a
h
o
o
.c
o
m
2
,
p
a
ta
n
a
v
i
j
i
t@y
a
h
o
o
.c
o
m
3
Ab
strac
t
B
e
c
a
u
s
e
o
f
t
h
e
f
e
a
s
i
b
l
e
a
n
d
i
m
p
r
e
s
s
i
v
e
f
a
l
l
o
u
t
,
t
h
e
c
l
a
s
s
i
c
a
l
S
u
p
e
r
-
R
e
s
o
l
u
t
i
o
n
R
e
c
o
n
s
t
r
u
c
t
i
o
n
(
S
R
R
)
i
s
t
h
e
c
o
n
t
e
m
p
o
r
a
r
y
a
l
g
o
r
i
t
h
m
f
o
r
i
m
p
r
o
v
i
n
g
s
p
a
t
i
a
l
i
n
f
o
r
m
a
t
i
o
n
a
n
d
r
e
d
u
c
i
n
g
n
o
i
s
e
a
n
d
S
I
S
R
(
S
i
n
g
l
e
-
I
m
a
g
e
S
u
p
e
r
-
R
e
s
o
l
u
t
i
o
n
)
m
e
t
h
o
d
,
w
h
i
c
h
i
s
f
o
r
m
o
n
t
h
e
c
l
a
s
s
i
c
a
l
S
R
R
,
i
s
s
o
l
e
l
y
d
e
v
e
l
o
p
e
d
f
o
r
i
m
p
r
o
v
i
n
g
s
p
a
t
i
a
l
i
n
f
o
r
m
a
t
i
o
n
.
D
i
s
a
s
t
r
o
u
s
l
y
,
d
e
f
i
c
i
e
n
c
y
o
f
t
h
e
c
l
a
s
s
i
c
a
l
S
I
S
R
m
e
t
h
o
d
i
s
c
o
n
c
e
p
t
u
a
l
l
y
c
o
m
p
u
t
e
d
f
r
o
m
t
h
r
e
e
s
p
e
c
i
f
i
c
a
t
i
o
n
s
(
b
,
h
,
k
)
a
n
d
t
h
e
s
i
m
u
l
a
t
i
n
g
c
a
l
c
u
l
a
t
i
o
n
o
f
t
h
e
o
p
t
i
m
i
z
e
d
s
p
e
c
i
f
i
c
a
t
i
o
n
s
f
o
r
i
n
t
e
r
p
o
l
a
t
i
n
g
t
h
e
b
e
t
t
e
r
a
n
d
h
i
g
h
e
r
s
p
a
t
i
a
l
i
n
f
o
r
m
a
t
i
o
n
i
m
a
g
e
s
w
i
t
h
h
i
g
h
e
s
t
P
S
N
R
i
s
s
o
b
u
r
d
e
r
s
o
m
e
.
F
o
r
f
i
g
u
r
i
n
g
o
u
t
t
h
i
s
i
s
s
u
e
,
t
h
e
G
e
m
a
n
&M
c
C
l
u
r
e
f
u
n
c
t
i
o
n
i
s
p
r
o
p
o
s
e
d
t
o
r
e
p
l
a
c
e
w
i
t
h
t
h
e
o
r
d
i
n
a
r
y
S
I
S
R
f
u
n
c
t
i
o
n
b
e
c
a
u
s
e
t
h
i
s
f
u
n
c
t
i
o
n
i
s
c
o
n
c
e
p
t
u
a
l
l
y
c
o
m
p
u
t
e
d
f
r
o
m
o
n
l
y
o
n
e
s
p
e
c
i
f
i
c
a
t
i
o
n
(
T
)
,
c
o
n
t
r
a
r
y
t
o
t
h
r
e
e
s
p
e
c
i
f
i
c
a
t
i
o
n
s
s
i
m
i
l
a
r
t
o
c
l
a
s
s
i
c
a
l
S
I
S
R
m
e
t
h
o
d
h
e
n
c
e
t
h
i
s
a
n
a
l
y
t
i
c
a
r
t
i
c
l
e
f
o
c
u
s
e
s
t
o
o
f
f
e
r
a
n
o
v
e
l
e
l
e
m
e
n
t
a
r
y
s
p
a
t
i
a
l
e
x
p
a
n
d
i
n
g
s
c
h
e
m
e
f
o
r
m
o
n
S
I
S
R
m
e
t
h
o
d
w
i
t
h
m
o
d
i
f
y
i
n
g
G
e
m
a
n
&M
c
C
l
u
r
e
f
u
n
c
t
i
o
n
.
T
h
e
r
e
f
o
r
e
,
t
h
e
f
a
l
l
o
u
t
o
f
a
p
r
o
p
o
s
e
d
s
p
a
t
i
a
l
e
x
p
a
n
d
i
n
g
s
c
h
e
m
e
a
p
p
r
o
x
i
m
a
t
e
l
y
m
a
t
c
h
e
s
t
o
c
l
a
s
s
i
c
a
l
S
I
S
R
m
e
t
h
o
d
.
F
r
o
m
t
h
e
s
e
r
e
a
s
o
n
,
a
n
o
v
e
l
e
l
e
m
e
n
t
a
r
y
s
p
a
t
i
a
l
e
x
p
a
n
d
i
n
g
s
c
h
e
m
e
i
s
e
a
s
i
l
y
i
m
p
l
e
m
e
n
t
e
d
f
o
r
r
e
a
l
w
o
r
k
s
.
Key
w
ords
:
d
i
g
i
t
a
l
i
m
a
g
e
i
n
te
r
p
o
l
a
t
i
o
n
,
d
i
g
i
ta
l
i
m
a
g
e
p
ro
c
e
s
s
i
n
g
,
G
e
m
a
n
&M
c
Clu
re
fu
n
c
ti
o
n
,
s
i
n
g
l
e
-
i
m
a
g
e
s
u
p
e
r
re
s
o
l
u
t
i
o
n
(SS
IR)
,
s
u
p
e
r
-
re
s
o
l
u
ti
o
n
re
c
o
n
s
tru
c
ti
o
n
(SS
R)
Copy
righ
t
©
2
0
1
9
Uni
v
e
rsi
t
a
s
Ahm
a
d
D
a
hl
a
n.
All
rig
ht
s
r
e
s
e
rve
d
.
1. C
o
mp
r
ement
ar
y I
n
v
es
t
i
g
atio
n
B
ec
au
s
e
of
great
r
eq
ui
r
em
en
t
of
c
on
te
m
po
r
ar
y
c
om
pu
ter
v
i
s
i
o
n
an
d
i
m
ag
e
proc
es
s
i
ng
[
1
-
5]
f
or
m
ore
tha
n
on
e
an
d
ha
l
f
de
c
ad
e
s
,
i
m
ag
es
wi
th
be
tt
er
an
d
hi
g
he
r
s
pa
ti
a
l
i
nf
orm
ati
on
i
s
dram
ati
c
al
l
y
r
eq
ui
r
e
d
ho
wev
er
al
m
os
t
r
ec
orded
i
m
ag
es
i
s
us
ua
l
l
y
l
o
wer
s
pa
t
i
al
i
nf
orm
ati
on
,
w
h
i
c
h
ar
e
r
ec
o
r
de
d
b
y
m
ark
eti
ng
bu
i
l
d
-
i
n
CCT
V
or
c
am
era.
T
he
r
ef
or
e,
th
e
r
eg
ul
ar
di
g
i
ta
l
i
m
ag
e
i
nt
erpo
l
at
i
o
n
i
s
c
on
c
ep
tua
l
i
d
ea
i
n
order
to
c
r
ea
ti
n
g
an
i
m
ag
e
wi
th
hi
gh
er
s
pa
ti
al
i
nf
orm
ati
on
f
r
o
m
on
e
i
m
ag
e
wi
th
l
o
wer
s
pa
ti
al
i
nf
orm
ati
o
n.
T
he
do
m
i
na
nt
as
p
i
r
a
ti
on
of
s
pa
ti
a
l
ex
pa
nd
i
n
g
s
c
he
m
e
f
or
m
o
n
S
I
S
R
m
eth
od
,
wh
i
c
h
i
s
de
v
el
op
e
d
f
r
om
the
w
or
l
d
w
i
d
e
f
am
ou
s
S
up
er
-
Res
ol
uti
on
R
ec
on
s
t
r
uc
ti
on
(
S
RR)
[
6
-
17
],
i
s
f
or
s
ol
e
l
y
i
m
prov
i
ng
s
p
ati
al
i
nf
orm
ati
on
.
Conc
ep
t
ua
l
l
y
,
b
oth
S
RR
a
n
d
S
I
S
R
are
an
i
l
l
-
po
s
ed
i
n
v
ers
ed
i
s
s
ue
[1
0
,
1
6
]
f
r
o
m
the
r
ea
s
o
n
tha
t
c
op
i
o
us
hi
g
he
r
s
pa
t
i
a
l
i
nf
orm
ati
on
i
m
ag
es
c
an
be
c
on
s
ti
tut
ed
f
r
om
on
l
y
on
e
r
ec
ord
ed
i
m
ag
e
w
i
th
l
o
w
er
s
pa
ti
a
l
i
nf
or
m
ati
on
u
n
de
r
th
e
i
d
en
t
i
c
al
r
ec
ord
i
n
g
proc
es
s
.
In
the
S
IS
R
i
n
qu
i
r
i
ng
i
n
th
e
l
as
t
two
a
nd
ha
l
f
de
c
ad
es
,
o
ne
of
the
m
os
t
f
ea
s
i
bl
e
a
nd
i
m
pres
s
i
v
e
s
pa
t
i
a
l
ex
p
an
d
i
n
g
s
c
he
m
e
i
s
th
e
S
IS
R
m
eth
od
us
i
ng
bo
un
d
ar
y
c
on
s
tr
uc
t
i
ng
tec
h
ni
q
ue
s
,
whi
c
h
i
s
or
i
g
i
na
l
l
y
i
n
v
en
t
ed
b
y
B
o
-
W
on
J
eo
n
[
18
]
i
n
20
0
6.
T
hi
s
m
et
ho
d
c
an
c
on
s
ti
t
ute
b
ett
er
a
nd
hi
g
he
r
s
pa
t
i
al
i
nf
orm
ati
on
i
m
ag
es
an
d,
m
oreov
er,
ha
s
l
o
w
c
om
pl
ex
i
t
y
.
T
hi
s
S
I
S
R
m
eth
od
i
s
i
n
c
orpor
ate
d
b
y
t
w
o
do
m
i
na
n
t
s
ub
-
s
y
s
tem
s
:
the
bo
u
nd
ar
y
c
on
s
tr
uc
t
i
ng
s
ub
-
s
y
s
t
em
an
d
th
e
r
eg
ul
a
r
i
z
e
d
f
un
c
ti
on
.
I
n
t
he
f
i
r
s
t
s
ub
-
s
y
s
tem
,
the
bo
un
da
r
y
c
o
ns
tr
uc
ti
n
g
s
ub
-
s
y
s
tem
i
s
c
on
c
ep
tua
l
l
y
c
om
pu
ted
f
r
om
La
pl
a
c
i
an
p
y
r
am
i
d
m
eth
od
[19
-
2
3]
f
or
c
on
s
tr
uc
ti
ng
bo
un
d
ar
y
an
d
l
i
ne
of
th
e
hi
gh
er
s
pa
t
i
al
i
nf
or
m
ati
on
i
m
ag
e.
In
the
s
ec
on
d
s
ub
-
s
y
s
te
m
,
the
r
eg
ul
ari
z
ed
f
un
c
ti
on
C
(x
,
y
)
[18
]
i
s
c
on
c
e
ptu
a
l
l
y
c
om
pu
ted
f
r
o
m
three
s
pe
c
i
f
i
c
ati
o
ns
(
b,
h,
k
)
an
d
the
s
i
m
ul
ati
ng
c
al
c
u
l
at
i
o
n
o
f
t
he
op
ti
m
i
z
e
d
s
pe
c
i
f
i
c
ati
o
ns
[24
,
2
5]
f
or
i
nte
r
p
ol
ati
ng
the
be
t
ter
a
nd
hi
gh
er
s
p
ati
al
i
nf
orm
ati
on
i
m
ag
es
wi
th
hi
gh
es
t
P
S
NR
.
F
or
f
i
gu
r
i
ng
ou
t
t
hi
s
i
s
s
u
e
,
th
e
G
em
an
&
Mc
Cl
ure
f
un
c
ti
on
[
26
],
whi
c
h
h
as
be
en
m
ath
em
ati
c
al
l
y
gu
ara
nte
e
an
d
i
nv
es
ti
g
ate
d
f
or
m
or
e
tha
n
t
hi
r
t
y
y
e
ars
the
r
eb
y
t
he
G
em
an
&
Mc
C
l
ure
f
un
c
ti
on
ha
s
an
ex
h
au
s
ti
v
e
i
n
qu
i
r
i
ng
,
i
s
c
on
c
ep
tu
al
l
y
c
om
pu
ted
f
r
om
on
l
y
o
ne
s
pe
c
i
f
i
c
ati
o
n
(
T
)
,
c
on
tr
ar
y
to
th
r
ee
s
pe
c
i
f
i
c
at
i
on
s
s
i
m
i
l
ar
to
c
l
as
s
i
c
al
S
I
S
R
m
eth
od
.
T
he
r
ef
ore,
thi
s
an
a
l
y
t
i
c
arti
c
l
e
f
oc
us
es
to
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
A
no
v
e
l
e
l
e
me
n
tary
s
pa
t
i
al
ex
pa
nd
i
n
g s
c
he
m
e f
orm
on
S
IS
R m
eth
o
d
... (
D
arun
K
e
s
r
arat
)
2555
of
f
er
a
no
v
el
el
em
en
tar
y
s
pa
ti
a
l
ex
p
an
d
i
n
g
s
c
he
m
e
f
or
m
on
S
IS
R
m
eth
od
a
nd
m
od
i
f
y
i
ng
G
em
an
&
Mc
Cl
ure
f
un
c
ti
on
.
T
he
an
al
y
ti
c
art
i
c
l
e
c
a
n
b
e
ad
j
us
ted
as
en
s
u
i
ng
.
Th
e
f
i
r
s
t
organ
i
z
ed
c
on
t
en
t
i
s
the
c
o
m
pl
em
en
tar
y
i
n
v
es
t
i
ga
ti
o
n
of
the
S
S
R
m
eth
od
an
d
S
IS
R
m
eth
od
.
Nex
t,
th
e
s
ec
on
d
orga
ni
z
e
d
c
on
ten
t
i
s
t
he
c
l
as
s
i
c
al
s
pa
t
i
al
ex
pa
nd
i
ng
s
c
he
m
e f
or
m
on
S
IS
R m
eth
od
.
S
u
bs
eq
u
en
tl
y
,
the
th
i
r
d
organ
i
z
ed
c
o
nte
nt
i
s
th
e
n
ov
e
l
s
p
ati
al
ex
pa
nd
i
ng
s
c
he
m
e
f
or
m
on
propos
ed
G
e
m
an
&
Mc
Cl
ure
S
IS
R
m
eth
od
.
La
ter,
t
he
f
orth
org
an
i
z
e
d
c
o
nte
nt
i
s
c
o
m
pu
ter
c
o
m
pu
tat
i
on
s
u
bs
ta
nti
a
ti
o
n,
whi
c
h
are
c
om
pa
r
ati
v
e
l
y
as
s
es
s
ed
to
thre
e
ord
i
na
r
y
s
pa
t
i
a
l
ex
pa
nd
i
n
g
s
c
he
m
es
:
ne
ares
t,
bi
l
i
n
ea
r
an
d
bi
c
ub
i
c
, o
n
bo
un
t
i
f
ul
i
m
ag
e
s
. Fi
na
l
l
y
, t
he
f
i
f
th
orga
ni
z
e
d c
on
te
nt
i
s
th
e
the
atri
c
a
l
c
on
c
i
s
i
o
n.
2.
T
h
e Cla
ss
ica
l
S
p
ati
al
E
xpan
d
ing
S
ch
eme
f
o
r
m
o
n
S
IS
R M
eth
o
d
T
hi
s
organi
z
e
d
c
on
ten
t
r
ap
i
dl
y
ex
p
ou
n
ds
the
a
l
go
r
i
thm
i
c
ap
preh
en
s
i
on
of
ordi
n
ar
y
S
IS
R
m
eth
od
[18
]
us
i
n
g
b
ou
n
d
ar
y
c
on
s
tr
uc
t
i
ng
tec
h
ni
q
u
es
,
w
h
i
c
h
i
s
c
om
pu
ted
f
r
om
bo
un
da
r
y
c
on
s
tr
uc
ti
ng
proc
ed
ure (
as
ex
hi
b
i
t
i
n F
i
gu
r
e
1) an
d
G
e
m
an
&
Mc
Cl
ure
r
e
gu
l
ari
z
at
i
o
n p
r
oc
e
d
ure.
2
.1.
B
o
u
n
d
a
r
y
C
o
n
struct
i
n
g
P
r
o
c
edu
r
e
T
he
v
i
ta
l
target
of
thi
s
bo
un
da
r
y
c
o
ns
tr
uc
ti
n
g
proc
ed
ure
[1
8
]
i
s
f
or
c
on
s
tr
uc
ti
ng
the
bo
u
nd
ar
y
,
de
tai
l
an
d
ed
ge
of
the
b
ett
er
an
d
hi
gh
er
s
pa
t
i
a
l
i
nf
orm
ati
on
i
m
ag
es
wi
th
hi
g
he
s
t P
S
NR
.
A
t
f
i
r
s
t,
the
r
ec
orde
d
l
o
w
s
pa
ti
al
i
nf
or
m
ati
on
i
m
ag
e
(
(
)
0
,
G
x
y
at
mn
)
s
pa
ti
al
l
y
s
c
al
es
do
w
n
to
c
on
s
tr
uc
ti
ng
a
c
l
a
s
s
of
l
o
w
s
pa
t
i
al
i
nf
orm
ati
on
i
m
ag
e
s
as
(
)
1
,
G
x
y
at
22
mn
,
(
)
2
,
G
x
y
at
44
mn
,
(
)
3
,
G
x
y
at
88
mn
and
(
)
4
,
G
x
y
at
16
16
mn
.
A
t
s
ec
on
d
,
op
erati
ng
a
ordi
n
ar
y
s
pa
t
i
a
l
ex
pa
nd
i
ng
s
c
he
m
e
,
a
c
l
as
s
of
m
ore
s
pa
ti
al
i
nf
orm
ati
on
i
m
ag
e
s
are
c
on
s
tr
uc
t
ed
f
r
om
ea
c
h
l
o
w
s
pa
ti
a
l
i
nf
orm
ati
on
i
m
ag
e
s
(
fr
o
m
f
i
r
s
t
proc
ed
ure)
as
(
)
0
,
L
x
y
at
mn
fr
om
(
)
1
,
G
x
y
at
22
mn
,
(
)
1
,
L
x
y
at
22
mn
f
r
o
m
(
)
2
,
G
x
y
at
44
mn
,
(
)
2
,
L
x
y
at
44
mn
f
r
o
m
(
)
3
,
G
x
y
at
88
mn
and
(
)
3
,
L
x
y
at
88
mn
f
r
o
m
(
)
4
,
G
x
y
at
16
16
mn
.
A
t
t
hi
r
d,
op
erati
ng
La
p
l
ac
i
a
n
hi
s
t
og
r
am
c
o
m
pu
tat
i
o
n
a
nd
G
au
s
s
i
an
s
ta
ti
s
ti
c
a
l
an
al
y
s
i
s
,
the
f
i
r
s
t
an
d
s
ec
on
d
m
o
m
en
t
s
tat
i
s
ti
c
f
al
l
o
ut
a
t
e
ac
h
s
pa
ti
al
l
e
v
e
l
of
the
bo
u
nd
ar
y
,
de
ta
i
l
an
d
ed
ge
i
m
ag
es
are
c
on
s
tr
uc
t
ed
as
(
)
00
,
at
mn
,
(
)
11
,
at
22
mn
,
(
)
22
,
at
44
mn
an
d
(
)
33
,
at
88
mn
.
A
t
f
ou
r
th,
f
or
r
ap
i
d
c
ou
nti
ng
an
d
r
am
i
f
i
c
ati
on
r
ed
uc
t
i
on
b
y
us
i
ng
L
ap
l
ac
i
a
n
s
tat
i
s
ti
c
a
l
an
a
l
y
s
i
s
,
the
f
i
r
s
t
an
d
s
ec
on
d
m
om
en
t
s
tat
i
s
ti
c
f
al
l
o
uts
(
,
)
are
r
ed
uc
ed
to
be
o
ne
s
tat
i
s
ti
c
f
al
l
ou
t
(
)
as
0
at
mn
,
1
at
22
mn
,
2
at
44
mn
an
d
3
at
44
mn
.
A
t
f
i
f
th,
b
y
op
era
ti
n
g
t
he
l
ea
s
t
s
qu
are
i
nte
r
po
r
at
i
ng
s
c
he
m
e,
the
b
ou
n
da
r
y
,
d
et
ai
l
an
d
ed
ge
i
m
ag
e o
f
th
e
hi
gh
er s
pa
ti
al
i
nf
orm
ati
on
i
m
ag
es
(
)
1
,
L
x
y
−
%
at
22
mn
is
r
ec
on
s
tr
uc
te
d.
2
.
2
.
O
r
d
in
ar
y R
egu
l
ar
iz
ati
o
n
P
r
o
c
edu
r
e
T
he
v
i
ta
l
targ
et
of
thi
s
b
ou
nd
ar
y
c
on
s
tr
uc
t
i
ng
proc
ed
ure
i
s
f
or
f
orc
ei
ng
the
up
pe
r
l
i
m
i
tat
i
o
n
of
the
de
tai
l
an
d
ed
g
e
i
m
ag
e
of
the
hi
g
h
er
s
pa
ti
a
l
i
nf
or
m
ati
on
(
(
)
,
C
x
y
)
tha
t
i
s
c
on
s
tr
uc
ted
b
y
the
b
ou
nd
a
r
y
c
o
ns
tr
uc
ti
n
g
proc
ed
ure
wi
th
t
he
or
d
i
na
r
y
r
eg
u
l
ar
i
z
a
ti
on
proc
ed
ure
as
en
s
ui
ng
:
(
)
(
)
(
)
(
)
(
)
(
)
(
)
,
,,
,
M
x
y
b
k
M
x
y
M
x
y
h
C
x
y
+
+
=
(
1)
where
b
,
k
an
d
h
are f
i
x
s
pe
c
i
f
i
c
ati
on
s
A
t
f
i
na
l
,
the
hi
g
h
s
pa
t
i
a
l
i
nf
or
m
ati
on
i
m
ag
e
c
an
be
c
on
s
tr
uc
t
ed
b
y
a
dd
ng
of
the
bo
un
da
r
y
,
d
eta
i
l
an
d
ed
ge
i
m
ag
e
of
the
hi
gh
er
s
p
ati
al
i
nf
or
m
ati
on
i
m
ag
es
,
w
h
i
c
h
are
m
ul
ti
y
p
l
e
d
wi
th
t
he
f
al
l
ou
t
of
the
r
e
g
ul
ari
z
a
ti
on
proc
e
du
r
e
(
(
)
(
)
(
)
11
ˆ
,
,
,
L
x
y
C
x
y
L
x
y
−−
=
%
)
an
d
the
hi
gh
er
s
pa
ti
a
l
i
nf
orm
ati
on
i
m
ag
e
(
(
)
1
,
−
%
G
x
y
)
at
22
mn
,
w
h
i
c
h
i
s
c
on
s
tr
uc
ted
b
y
op
erat
i
ng
a
ordi
na
r
y
s
pa
ti
a
l
ex
p
an
d
i
n
g s
c
he
m
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
5,
O
c
tob
er 20
19
:
25
5
4
-
25
60
2556
3.
T
he
N
o
v
el
E
le
men
t
ar
y
S
p
atial
E
xpan
d
ing
S
ch
em
e
f
o
r
m
o
n
S
IS
R
M
eth
o
d
w
it
h
M
o
d
if
y
ing
G
eman&
M
cC
l
u
r
e
F
u
n
ctio
n
A
s
ex
po
u
nd
s
i
n
th
e
p
r
ev
ou
s
orgi
an
z
i
e
d
c
on
t
e
nt
,
th
e
ord
i
na
r
y
r
eg
ul
ar
i
z
ati
on
proc
ed
ure
[1
8
]
,
whi
c
h
i
s
de
f
i
ne
d
i
n
(
1),
c
on
c
ep
t
ua
l
l
y
c
o
m
pu
ted
f
r
o
m
three
s
pe
c
i
f
i
c
ati
o
ns
(
b,
h,
k
)
the
r
eb
y
th
e
s
i
m
ul
ati
n
g
c
al
c
ul
at
i
o
n
of
the
op
t
i
m
i
z
ed
s
p
ec
i
f
i
c
ati
o
ns
f
or
i
nte
r
po
l
ati
ng
the
be
tt
er
an
d
hi
g
he
r
s
pa
t
i
a
l
i
nf
orm
ati
on
i
m
ag
es
w
i
t
h
h
i
gh
es
t
P
S
NR
i
s
s
o
b
urder
s
om
e.
F
or
f
i
gu
r
i
ng
ou
t
th
i
s
i
s
s
ue
,
t
he
G
em
an
&
Mc
C
l
ure
f
un
c
ti
o
n
[
26
]
i
s
c
o
nc
ep
tu
al
l
y
c
om
pu
ted
f
r
om
on
l
y
o
ne
s
pe
c
i
f
i
c
ati
o
n (T
)
, c
on
tr
ar
y
t
o t
hree
s
pe
c
i
f
i
c
at
i
on
s
s
i
m
i
l
a
r
to
c
l
as
s
i
c
al
S
I
S
R m
eth
od
.
T
he
G
em
an
&
Mc
Cl
ure
f
un
c
ti
on
an
d
i
ts
grad
i
en
t
i
s
de
f
i
ne
d
as
e
ns
ui
ng
(
2
)
a
nd
(
3
)
,
r
es
pe
c
ti
v
el
y
:
(
)
(
)
(
)
(
)
(
)
2
2
2
2
,
,
,
G
M
G
M
GM
M
x
y
C
x
y
T
T
M
x
y
=
+
(
2
)
(
)
(
)
(
)
(
)
(
)
4
2
2
4
2,
,
,
G
M
G
M
GM
M
x
y
C
x
y
T
T
M
x
y
=
+
(
3
)
where
GM
T
i
s
G
em
an
&
Mc
Cl
ure
s
pe
c
i
f
i
c
ati
o
n,
w
h
i
c
h
i
s
c
om
pu
te
d b
y
f
ul
l
s
ea
r
c
h m
eth
od
.
4.
C
o
mp
u
t
er
C
o
mp
u
t
atio
n
S
u
b
stat
n
t
iatio
n
T
hi
s
organi
z
ed
c
on
ten
t
ex
ha
us
ti
v
e
l
y
ex
po
u
nd
s
s
ub
s
tan
ti
ati
on
f
r
om
c
om
pu
ter
c
o
m
pu
tat
i
on
,
w
h
i
c
h a
r
e
c
ov
ered b
ou
nti
f
ul
i
m
ag
es
s
uc
h
as
Cam
era
m
an
, T
r
ee
, Hou
s
e,
Res
o
l
ut
i
on
c
ha
r
t,
S
tr
e
am
an
d
brid
ge
,
S
ai
l
b
oa
t,
P
ep
pe
r
s
,
T
i
ff
an
y
,
B
ab
o
on
,
L
en
a,
F
-
16
,
M
ob
i
l
e
F
r
am
e1
0,
P
en
t
ag
o
n,
A
er
i
a
l
[
24
,
25
]
as
ex
po
un
d
i
n
T
ab
l
e
1
a
n
d
are
c
om
pu
ti
ng
b
y
MA
T
LA
B
s
of
t
w
are.
A
t
f
i
r
s
t,
ea
c
h
ord
i
na
r
y
i
m
ag
e
i
s
s
c
a
l
ed
d
o
w
n
a
nd
c
o
m
bi
ne
d
wi
th
no
i
s
e
f
or
r
ec
on
s
tr
uc
ti
on
the
r
ec
ord
ed
i
m
ag
e.
N
ex
t,
the
no
v
e
l
e
l
em
en
tar
y
s
pa
t
i
al
ex
pa
n
di
ng
S
I
S
R
s
c
he
m
e,
th
e
ordi
n
ar
y
s
pa
ti
a
l
ex
pa
n
d
i
ng
S
IS
R
s
c
he
m
e,
s
pa
ti
al
ex
p
an
di
n
g
n
ea
r
es
t
s
c
he
m
e,
s
pa
ti
al
ex
p
an
d
i
ng
b
i
l
i
n
ea
r
s
c
he
m
e
an
d
s
pa
t
i
a
l
ex
p
an
d
i
ng
bi
c
ub
i
c
s
c
he
m
e
are
op
er
ate
d
on
e
ac
h
r
ec
orded
i
m
ag
e
f
or
c
o
m
pa
r
ati
v
e
l
y
as
s
es
s
i
ng
th
e
f
ea
s
i
bl
e
a
nd
i
m
pres
s
i
v
e
f
al
l
ou
t
of
the
pro
po
s
ed
s
pa
ti
a
l
ex
pa
nd
i
n
g s
c
he
m
e.
T
ab
l
e 1
. T
he
C
om
pu
ter Co
m
pu
tat
i
on
S
ub
s
t
an
t
i
at
i
on
I
m
a
g
e
N
a
m
e
R
e
s
o
lut
ion
I
m
a
g
e
N
a
m
e
R
e
s
o
lut
ion
C
a
m
e
r
a
m
a
n
2
5
6
x
2
5
6
Tif
f
a
n
y
5
1
2
x
5
1
2
Tr
e
e
2
5
6
x
2
5
6
B
a
b
o
o
n
5
1
2
x
5
1
2
H
o
u
s
e
2
5
6
x
2
5
6
Lena
5
1
2
x
5
1
2
R
e
s
o
lut
ion
c
h
a
r
t
2
5
6
x
2
5
6
F
-
16
5
1
2
x
5
1
2
S
t
e
a
m
&
b
r
idg
e
5
1
2
x
5
1
2
M
o
b
il
e
Fr
a
m
e
10
7
0
4
x
4
8
0
S
a
il
b
o
a
t
5
1
2
x
5
1
2
P
e
n
t
a
g
o
n
1
0
2
4
x
1
0
2
4
P
e
p
p
e
r
s
5
1
2
x
5
1
2
A
e
r
ial
1
0
2
4
x
1
0
2
4
4
.
1
.
No
O
u
t
li
er
A
t
f
i
r
s
t,
th
e
org
an
i
z
ed
s
u
b
-
c
on
ten
t
ex
po
un
ds
the
op
ti
m
i
z
at
i
o
n
of
G
em
an
&
M
c
Cl
ure
s
pe
c
i
f
i
c
ati
o
n (
GM
T
)
th
at
i
s
m
od
i
f
i
ed
du
r
i
ng
0 t
o 2
0
00
f
or i
nte
r
po
l
at
i
n
g t
h
e b
ett
er
an
d
h
i
g
he
r
s
pa
t
i
a
l
i
nf
orm
ati
on
i
m
ag
es
w
i
t
h
hi
g
he
s
t
P
S
NR.
T
he
c
o
m
pu
ter
c
o
m
pu
tat
i
on
s
ub
s
ta
nti
ati
on
of
no
ou
tl
i
er
i
s
ex
po
un
de
d
i
n
th
e
F
i
gu
r
e
2.
F
r
om
thi
s
s
ub
s
tan
t
i
at
i
o
n
i
n
th
i
s
f
i
gu
r
e,
i
t
c
a
n
be
f
i
r
m
ed
tha
t
th
e
G
em
an
&
Mc
Cl
ure
s
pe
c
i
f
i
c
a
ti
on
i
s
f
i
x
to
be
o
pti
m
i
z
e
d
at
9
00
-
10
50
f
or
no
r
m
al
i
m
ag
es
an
d
th
e
G
em
an
&
Mc
Cl
ure
s
p
ec
i
f
i
c
a
ti
on
i
s
f
i
x
to
b
e
o
pti
m
i
z
e
d
at
1
25
0
-
14
5
0
f
or
i
m
ag
e
wi
th
bo
u
nti
f
ul
bo
un
da
r
y
,
de
t
ai
l
an
d
e
dg
e
i
nf
orm
ati
on
.
T
he
gl
ob
a
l
c
o
m
pu
tat
i
o
n
s
ub
s
tan
ti
a
ti
o
n
i
n
P
S
NR
i
s
ex
po
un
de
d
i
n
T
ab
l
e
2
f
or
no
ou
tl
i
er.
T
he
f
ea
s
i
bl
e
an
d
i
m
pres
s
i
v
e
f
al
l
o
ut
of
the
no
v
e
l
el
em
en
tar
y
s
pa
ti
a
l
ex
p
an
d
i
ng
s
c
he
m
e
ap
prox
i
m
ate
l
y
m
atc
he
s
to
c
l
as
s
i
c
al
S
IS
R
m
eth
od
ne
v
erth
el
es
s
th
e
s
i
m
ul
ati
ng
c
a
l
c
ul
ati
on
of
the
op
t
i
m
i
z
ed
s
p
ec
i
f
i
c
ati
o
n
i
s
s
o
r
ea
di
l
y
an
d
r
a
pi
dl
y
c
o
m
pa
r
ed
w
i
t
h
c
l
as
s
i
c
al
S
IS
R m
eth
od
(
w
i
t
h t
hree
s
pe
c
i
f
i
c
at
i
on
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
A
no
v
e
l
e
l
e
me
n
tary
s
pa
t
i
al
ex
pa
nd
i
n
g s
c
he
m
e f
orm
on
S
IS
R m
eth
o
d
... (
D
arun
K
e
s
r
arat
)
2557
F
i
gu
r
e
1.
A
no
v
e
l
el
em
en
ta
r
y
s
pa
t
i
al
ex
pa
nd
i
ng
s
c
he
m
e f
or
m
on
S
I
S
R m
eth
od
wi
t
h b
o
un
d
ar
y
F
i
gu
r
e
2
. T
he
c
om
pu
ter c
om
pu
tat
i
on
s
u
bs
tan
t
i
at
i
on
of
th
e
G
em
an
&
Mc
C
l
ure
s
p
ec
i
f
i
c
ati
on
a
nd
the
f
ea
s
i
b
l
e f
al
l
o
ut
of
a
n
ov
el
e
l
em
en
tar
y
s
p
ati
al
ex
pa
n
di
n
g s
c
he
m
e
at
no
ou
t
l
i
er
T
ab
l
e
2
. T
he
G
l
ob
a
l
C
om
p
uta
t
i
on
S
u
bs
tan
t
i
at
i
o
n i
n P
S
NR of
N
o
O
utl
i
er
P
S
N
R
(
in
d
B
)
C
las
s
i
c
a
l
S
p
a
t
ial
E
x
p
a
n
d
ing
S
c
h
e
m
e
S
p
a
t
ial
E
x
p
a
n
d
ing
S
c
h
e
m
e
U
s
ing
S
I
S
R
n
e
a
r
e
s
t
b
il
ine
a
r
b
ic
u
b
ic
Or
d
ina
r
y
G&M
L
e
n
a
2
5
6
3
0
.
7
8
4
7
3
1
.
0
3
0
5
3
1
.
8
2
8
1
3
2
.
0
7
5
9
3
2
.
1
1
6
0
Gir
l
2
5
6
3
0
.
9
3
5
4
3
1
.
4
6
0
2
3
1
.
9
9
9
7
3
2
.
7
2
6
8
3
2
.
4
4
0
5
B
a
b
o
o
n
2
5
6
2
3
.
2
5
5
6
2
3
.
0
5
4
6
2
3
.
6
1
4
2
2
3
.
8
2
1
5
2
3
.
8
2
2
7
H
o
u
s
e
1
2
8
2
9
.
5
0
5
3
3
0
.
1
1
3
5
3
0
.
8
9
4
4
3
1
.
0
9
7
5
3
1
.
3
1
1
9
R
e
s
o
lut
ion
c
h
a
r
t
1
9
.
5
6
4
3
2
0
.
0
1
4
3
2
0
.
9
6
9
9
2
1
.
4
6
7
6
2
1
.
1
3
2
0
P
e
p
p
e
r
s
2
5
6
3
0
.
9
2
5
4
3
1
.
7
8
3
5
3
2
.
6
8
1
4
3
2
.
9
6
2
3
3
3
.
4
5
6
8
A
irp
lan
e
2
5
6
3
0
.
2
8
6
1
3
1
.
0
8
0
2
3
1
.
8
5
1
5
3
2
.
7
5
9
4
3
2
.
6
1
6
0
P
e
n
t
a
g
o
n
5
1
2
2
8
.
8
0
3
9
2
8
.
8
5
6
9
2
9
.
6
3
4
5
2
9
.
8
9
8
7
2
9
.
9
7
0
7
A
e
r
ial
5
1
2
2
5
.
6
7
3
0
2
6
.
2
0
2
5
2
7
.
4
8
3
9
2
7
.
8
9
4
4
2
7
.
9
8
1
5
Tr
e
e
1
2
8
2
5
.
5
0
5
8
2
6
.
2
1
2
1
2
7
.
3
2
0
1
2
7
.
9
0
7
7
2
7
.
8
7
7
1
S
a
il
b
o
a
t
o
n
la
k
e
2
8
.
7
8
8
1
2
9
.
3
0
0
6
3
0
.
3
9
9
0
3
0
.
7
1
7
0
3
0
.
8
2
2
7
C
a
m
e
r
a
m
a
n
2
5
.
4
8
8
4
2
5
.
6
5
9
0
2
6
.
5
2
6
7
2
6
.
8
7
3
5
2
6
.
9
4
7
2
S
t
e
a
m
&
b
r
idg
e
2
5
.
6
7
8
7
2
5
.
8
5
3
3
2
6
.
6
5
3
5
2
6
.
9
2
0
2
2
6
.
9
6
5
5
M
o
b
il
c
a
l
Fr
a
m
e
2
2
.
0
8
6
3
2
2
.
1
2
2
3
2
3
.
0
3
3
3
2
3
.
3
3
2
4
2
3
.
4
3
3
4
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
5,
O
c
tob
er 20
19
:
25
5
4
-
25
60
2558
4
.
2
.
G
au
ss
ian
O
u
t
lie
r
A
t
f
i
r
s
t,
f
or
G
au
s
s
i
an
ou
tl
i
e
r
at
35
dB
,
th
e
v
i
tal
t
arget
of
thi
s
c
o
m
pu
tat
i
o
n
s
ub
s
ta
nti
a
ti
o
n
i
s
the
c
a
l
c
ul
ati
on
of
th
e
op
ti
m
i
z
e
d
s
pe
c
i
f
i
c
at
i
on
s
whi
c
h
i
s
m
od
i
f
i
ed
du
r
i
ng
0
to
20
00
f
or
i
nte
r
p
ol
ati
ng
th
e
be
tt
er
an
d
hi
gh
er
s
pa
ti
al
i
nf
or
m
ati
on
i
m
ag
es
w
i
th
h
i
gh
e
s
t
P
S
NR
as
ex
po
un
de
d
i
n
th
e
F
i
gu
r
e
3
.
F
r
om
thi
s
s
ub
s
tan
t
i
at
i
o
n
i
n
th
i
s
f
i
gu
r
e,
i
t
c
an
be
f
i
r
m
ed
tha
t
th
e
G
e
m
an
&
Mc
Cl
ure
s
pe
c
i
f
i
c
ati
o
n
i
s
f
i
x
to
be
o
pti
m
i
z
ed
at
1
05
0
-
1
25
0
f
or
no
r
m
al
i
m
ag
es
an
d
the
G
em
an
&
Mc
Cl
ure
s
pe
c
i
f
i
c
ati
o
n
i
s
f
i
x
to
be
op
ti
m
i
z
ed
at
14
0
0
-
16
0
0
f
or
i
m
ag
e
w
i
t
h
bo
u
nti
f
ul
bo
u
nd
ar
y
,
de
ta
i
l
a
nd
ed
ge
i
nf
orm
ati
on
.
T
he
g
l
ob
al
c
om
pu
tat
i
on
s
ub
s
t
an
t
i
at
i
on
i
n
P
S
NR
i
s
ex
po
un
de
d
i
n
T
ab
l
e
3
f
or
35
d
B
G
a
us
s
i
an
ou
t
l
i
er
.
T
he
f
ea
s
i
b
l
e
an
d
i
m
pres
s
i
v
e
f
al
l
ou
t
of
the
no
v
e
l
el
e
m
en
tar
y
s
p
ati
al
ex
pa
nd
i
n
g
s
c
he
m
e
ap
pro
x
i
m
ate
l
y
m
atc
he
s
to
c
l
a
s
s
i
c
al
S
IS
R
m
eth
od
ne
v
erthel
es
s
th
e
s
i
m
ul
ati
ng
c
a
l
c
ul
ati
on
of
the
op
t
i
m
i
z
ed
s
p
ec
i
f
i
c
ati
o
n
i
s
s
o
r
ea
di
l
y
an
d
r
a
pi
dl
y
c
o
m
pa
r
ed
w
i
t
h
c
l
as
s
i
c
al
S
IS
R m
eth
od
(
w
i
t
h t
hree
s
pe
c
i
f
i
c
at
i
on
)
.
F
i
gu
r
e
3
. T
he
c
om
pu
ter c
om
pu
tat
i
on
s
u
bs
tan
t
i
at
i
on
of
th
e
G
em
an
&
Mc
C
l
ure
s
p
ec
i
f
i
c
ati
on
a
nd
the
f
ea
s
i
b
l
e f
al
l
o
ut
of
a
n
ov
el
e
l
em
en
tar
y
s
p
ati
al
ex
pa
n
di
n
g s
c
he
m
e
at
35
d
B
G
a
us
s
i
an
o
utl
i
er
T
ab
l
e
3
. T
he
G
l
ob
a
l
C
om
p
uta
t
i
on
S
u
bs
tan
t
i
at
i
o
n i
n P
S
NR of
35
dB
G
au
s
s
i
a
n
O
ut
l
i
er
P
S
N
R
(
in
d
B
)
C
las
s
i
c
a
l
S
p
a
t
ial
E
x
p
a
n
d
ing
S
c
h
e
m
e
S
p
a
t
ial
E
x
p
a
n
d
ing
S
c
h
e
m
e
U
s
ing
S
I
S
R
n
e
a
r
e
s
t
b
il
ine
a
r
b
ic
u
b
ic
Or
d
ina
r
y
G&M
L
e
n
a
2
5
6
3
0
.
7
8
4
7
3
1
.
0
3
0
5
3
1
.
8
2
8
1
3
1
.
7
7
6
6
3
1
.
7
9
9
8
Gir
l
2
5
6
3
0
.
9
3
5
4
3
1
.
4
6
0
2
3
1
.
9
9
9
7
3
1
.
6
2
8
8
3
1
.
3
8
1
1
B
a
b
o
o
n
2
5
6
2
3
.
2
5
5
6
2
3
.
0
5
4
6
2
3
.
6
1
4
2
2
3
.
7
3
2
4
2
3
.
7
2
8
4
H
o
u
s
e
1
2
8
2
9
.
5
0
5
3
3
0
.
1
1
3
5
3
0
.
8
9
4
4
3
0
.
6
7
6
6
3
0
.
8
6
2
1
R
e
s
o
lut
ion
c
h
a
r
t
1
9
.
5
6
4
3
2
0
.
0
1
4
3
2
0
.
9
6
9
9
2
1
.
3
7
0
2
2
1
.
0
8
3
3
P
e
p
p
e
r
s
2
5
6
3
0
.
9
2
5
4
3
1
.
7
8
3
5
3
2
.
6
8
1
4
3
2
.
2
5
1
2
3
2
.
6
7
0
5
A
irp
lan
e
2
5
6
3
0
.
2
8
6
1
3
1
.
0
8
0
2
3
1
.
8
5
1
5
3
1
.
7
9
8
9
3
1
.
6
7
7
5
P
e
n
t
a
g
o
n
5
1
2
2
8
.
8
0
3
9
2
8
.
8
5
6
9
2
9
.
6
3
4
5
2
9
.
5
9
1
9
2
9
.
6
3
5
6
A
e
r
ial
5
1
2
2
5
.
6
7
3
0
2
6
.
2
0
2
5
2
7
.
4
8
3
9
2
7
.
5
5
6
5
2
7
.
6
3
4
8
Tr
e
e
1
2
8
2
5
.
5
0
5
8
2
6
.
2
1
2
1
2
7
.
3
2
0
1
2
7
.
6
9
8
1
2
7
.
6
6
1
7
S
a
il
b
o
a
t
o
n
la
k
e
2
8
.
7
8
8
1
2
9
.
3
0
0
6
3
0
.
3
9
9
0
3
0
.
3
5
5
2
3
0
.
4
5
6
3
C
a
m
e
r
a
m
a
n
2
5
.
4
8
8
4
2
5
.
6
5
9
0
2
6
.
5
2
6
7
2
6
.
7
3
7
8
2
6
.
8
0
3
8
S
t
e
a
m
&
b
r
idg
e
2
5
.
6
7
8
7
2
5
.
8
5
3
3
2
6
.
6
5
3
5
2
6
.
7
7
8
7
2
6
.
8
2
5
9
M
o
b
il
c
a
l
Fr
a
m
e
2
2
.
0
8
6
3
2
2
.
1
2
2
3
2
3
.
0
3
3
3
2
3
.
2
6
4
6
2
3
.
3
6
3
0
A
t
s
ec
on
d,
f
or
G
au
s
s
i
an
ou
tl
i
er
a
t
30
dB
,
th
e
v
i
t
al
t
arget
of
thi
s
c
om
pu
tat
i
on
s
ub
s
tan
t
i
at
i
on
i
s
the
c
al
c
ul
at
i
o
n
of
t
he
op
t
i
m
i
z
ed
s
pe
c
i
f
i
c
ati
o
ns
w
h
i
c
h
i
s
m
od
i
f
i
e
d
d
urin
g
0
to
20
0
0
f
or
i
nt
erpo
l
at
i
ng
the
b
ett
er
an
d
h
i
gh
er
s
pa
t
i
al
i
nf
orm
ati
on
i
m
ag
es
w
i
t
h
hi
g
he
s
t
P
S
N
R
as
ex
po
un
d
ed
i
n
the
F
i
gu
r
e
4.
F
r
om
thi
s
s
ub
s
tan
ti
ati
on
i
n
th
i
s
f
i
gu
r
e,
i
t
c
an
be
f
i
r
m
ed
tha
t
the
G
em
an
&
Mc
Cl
ure
s
pe
c
i
f
i
c
a
ti
on
i
s
f
i
x
to
be
op
t
i
m
i
z
e
d
at
1
60
0
-
1
65
0
f
or
no
r
m
al
i
m
ag
es
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
A
no
v
e
l
e
l
e
me
n
tary
s
pa
t
i
al
ex
pa
nd
i
n
g s
c
he
m
e f
orm
on
S
IS
R m
eth
o
d
... (
D
arun
K
e
s
r
arat
)
2559
the
G
em
an
&
Mc
Cl
ure
s
pe
c
i
f
i
c
ati
on
i
s
f
i
x
to
be
o
pti
m
i
z
ed
at
1
90
0
-
2
00
0
f
or
i
m
ag
e
wi
th
b
ou
nt
i
f
ul
bo
un
da
r
y
,
de
t
ai
l
an
d
e
dg
e
i
nf
orm
ati
on
.
T
he
gl
ob
a
l
c
o
m
pu
tat
i
o
n
s
ub
s
tan
ti
a
ti
o
n
i
n
P
S
NR
i
s
ex
po
un
de
d
i
n
T
ab
l
e
4
f
or
30
d
B
G
a
us
s
i
an
ou
t
l
i
er
.
T
he
f
ea
s
i
b
l
e
an
d
i
m
pres
s
i
v
e
f
al
l
o
ut
of
the
no
v
el
e
l
em
en
tar
y
s
pa
t
i
a
l
e
x
pa
nd
i
ng
s
c
he
m
e
ap
prox
i
m
ate
l
y
m
atc
he
s
to
c
l
as
s
i
c
al
S
I
S
R
m
eth
od
ne
v
erthe
l
es
s
th
e
s
i
m
ul
ati
n
g
c
al
c
u
l
at
i
on
of
the
o
pti
m
i
z
ed
s
pe
c
i
f
i
c
ati
on
i
s
s
o
r
ea
di
l
y
an
d
r
a
pi
dl
y
c
o
m
pa
r
ed
w
i
th
c
l
as
s
i
c
al
S
I
S
R m
eth
od
(
wi
th
t
hree s
p
e
c
i
f
i
c
ati
on
)
.
F
i
gu
r
e
4
. T
he
c
om
pu
ter c
om
pu
tat
i
on
s
u
bs
tan
t
i
at
i
on
of
th
e
G
em
an
&
Mc
C
l
ure
s
p
ec
i
f
i
c
ati
on
a
nd
the
f
ea
s
i
b
l
e f
al
l
o
ut
of
a
n
ov
el
e
l
em
en
tar
y
s
p
ati
al
ex
pa
n
di
n
g s
c
he
m
e
at
30
d
B
G
a
us
s
i
an
o
utl
i
er
T
ab
l
e
4
. T
he
G
l
ob
a
l
C
om
p
uta
t
i
on
S
u
bs
tan
t
i
at
i
o
n i
n P
S
NR of
30
dB
G
au
s
s
i
a
n
O
ut
l
i
er
P
S
N
R
(
in
d
B
)
C
las
s
i
c
a
l
S
p
a
t
ial
E
x
p
a
n
d
ing
S
c
h
e
m
e
S
p
a
t
ial
E
x
p
a
n
d
ing
S
c
h
e
m
e
U
s
ing
S
I
S
R
n
e
a
r
e
s
t
b
il
ine
a
r
b
ic
u
b
ic
Or
d
ina
r
y
G&M
L
e
n
a
2
5
6
2
9
.
8
8
3
4
3
0
.
6
2
2
8
3
1
.
0
5
5
8
3
1
.
1
8
5
9
3
1
.
1
8
5
8
Gir
l
2
5
6
2
8
.
3
6
5
0
2
9
.
9
1
0
0
2
9
.
6
3
8
2
2
9
.
7
1
2
0
2
9
.
6
8
6
9
B
a
b
o
o
n
2
5
6
2
2
.
9
5
6
0
2
2
.
9
3
9
4
2
3
.
3
9
8
4
2
3
.
5
3
5
9
2
3
.
5
3
6
0
H
o
u
s
e
1
2
8
2
8
.
3
2
1
9
2
9
.
5
5
6
8
2
9
.
8
3
0
2
3
0
.
0
0
5
5
3
0
.
0
0
6
3
R
e
s
o
lut
ion
c
h
a
r
t
1
9
.
3
6
0
9
1
9
.
8
0
7
3
2
0
.
7
0
9
1
2
1
.
2
3
8
6
2
0
.
9
1
5
5
P
e
p
p
e
r
s
2
5
6
2
9
.
2
4
8
0
3
0
.
8
8
9
9
3
1
.
0
2
3
6
3
1
.
3
1
1
2
3
1
.
3
1
3
4
A
irp
lan
e
2
5
6
2
8
.
2
0
3
6
2
9
.
9
5
7
0
2
9
.
8
5
6
5
3
0
.
0
9
7
5
3
0
.
0
9
5
5
P
e
n
t
a
g
o
n
5
1
2
2
7
.
8
8
5
8
2
8
.
4
5
8
6
2
8
.
8
8
2
1
2
9
.
0
0
0
9
2
9
.
0
0
0
8
A
e
r
ial
5
1
2
2
4
.
8
9
9
3
2
5
.
8
3
7
2
2
6
.
7
1
2
7
2
6
.
9
6
8
0
2
6
.
9
4
0
5
Tr
e
e
1
2
8
2
5
.
0
3
5
7
2
5
.
9
7
8
9
2
6
.
8
4
3
5
2
7
.
2
1
9
1
2
7
.
2
1
9
2
S
a
il
b
o
a
t
o
n
la
k
e
2
7
.
9
0
7
7
2
8
.
8
8
4
0
2
9
.
5
5
6
2
2
9
.
7
5
6
5
2
9
.
7
5
6
1
C
a
m
e
r
a
m
a
n
2
5
.
0
7
9
7
2
5
.
4
8
6
9
2
6
.
1
8
0
8
2
6
.
5
1
0
4
2
6
.
4
9
5
2
S
t
e
a
m
&
b
r
idg
e
2
5
.
3
0
0
3
2
5
.
6
8
8
0
2
6
.
3
3
5
1
2
6
.
5
4
6
6
2
6
.
5
4
6
5
M
o
b
il
c
a
l
Fr
a
m
e
2
1
.
8
9
1
4
2
2
.
0
4
7
0
2
2
.
8
7
5
8
2
3
.
2
1
4
4
2
3
.
2
1
3
0
5.
C
o
n
clus
ion
In
th
i
s
an
a
l
y
ti
c
as
s
es
s
i
n
g,
t
hi
s
an
al
y
t
i
c
arti
c
l
e
f
oc
us
es
to
of
f
er
a
no
v
e
l
e
l
em
en
tar
y
s
pa
ti
a
l
ex
pa
nd
i
n
g
s
c
he
m
e
f
or
m
on
S
I
S
R
m
eth
od
wi
th
m
od
i
f
y
i
ng
G
em
an
&
Mc
Cl
ure
f
un
c
ti
o
n
tha
t
i
s
c
on
c
ep
tu
al
l
y
c
om
pu
ted
f
r
om
on
l
y
on
e
s
pe
c
i
f
i
c
ati
on
(
T
)
,
c
on
tr
ar
y
t
o
three
s
pe
c
i
f
i
c
ati
o
ns
(
b,
h,
k
)
s
i
m
i
l
ar
to
c
l
as
s
i
c
al
S
IS
R
m
eth
od
.
B
y
ex
am
i
ni
ng
on
bo
un
ti
f
ul
i
m
ag
es
,
w
h
i
c
h
a
r
e
de
ba
s
ed
b
y
c
op
i
o
us
ou
tl
i
er
pa
tte
r
ns
,
i
n
an
a
l
y
t
i
c
a
l
ob
s
erv
ati
on
s
ec
ti
on
,
the
i
m
pres
s
i
v
e
f
al
l
ou
t
of
a
no
v
el
el
em
en
tar
y
s
pa
t
i
a
l
ex
p
an
di
n
g
s
c
he
m
e
ap
prox
i
m
ate
l
y
m
at
c
he
s
to
c
l
as
s
i
c
al
S
I
S
R
m
eth
od
ne
v
erthe
l
es
s
the
s
i
m
ul
ati
ng
c
al
c
ul
at
i
on
of
the
o
pti
m
i
z
e
d
s
pe
c
i
f
i
c
ati
o
n
(
whi
c
h
i
s
ba
s
ed
on
2x
10
3
s
ea
r
c
h
po
i
nts
)
i
s
s
o
r
ea
di
l
y
an
d
r
ap
i
d
l
y
c
om
pa
r
ed
wi
t
h
c
l
as
s
i
c
al
S
IS
R
m
eth
od
(
whi
c
h
i
s
ba
s
ed
on
8x
1
0
9
s
e
arc
h
po
i
nts
)
.
F
r
om
the
s
e
r
ea
s
on
,
a
n
ov
el
el
em
en
tar
y
s
pa
t
i
a
l
ex
p
an
d
i
ng
s
c
he
m
e
i
s
ea
s
i
l
y
i
m
pl
em
en
ted
f
or r
ea
l
wor
k
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
5,
O
c
tob
er 20
19
:
25
5
4
-
25
60
2560
A
c
kno
w
ledg
ement
s
T
he
r
es
ea
r
c
h p
r
oj
ec
t
w
as
f
un
de
d b
y
A
s
s
um
pti
on
Un
i
v
ers
i
t
y
.
Ref
er
en
ce
s
[1
]
J
a
i
n
AK.
Fu
n
d
a
m
e
n
t
a
l
s
o
f
Di
g
i
ta
l
I
m
a
g
e
Pro
c
e
s
s
i
n
g
.
En
g
l
e
w
o
o
d
Cl
i
ff
s
:
Pre
n
t
i
c
e
-
Hal
l
.
1
9
8
9
.
[2
]
G
o
n
z
a
l
e
z
RC
,
W
o
o
d
s
RE. Di
g
i
ta
l
I
m
a
g
e
Pro
c
e
s
s
i
n
g
.
2
n
d
e
d
.
Upp
e
r Sa
d
d
l
e
R
i
v
e
r:
Pre
n
ti
c
e
-
Hal
l
.
2
0
0
2
.
[3
]
Ab
d
u
l
w
a
h
e
d
MN
,
Ah
m
e
d
AK
,
Und
e
rw
a
te
r
Im
a
g
e
De
-
n
o
s
i
n
g
u
s
i
n
g
Di
s
c
r
e
te
W
a
v
e
l
e
t
T
ra
n
s
fo
r
m
a
n
d
Pre
-
w
h
i
te
n
i
n
g
Fi
l
te
r
.
TE
L
KO
M
NIKA
Te
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
C
o
m
p
u
ti
n
g
El
e
c
tro
n
i
c
s
a
n
d
Co
n
tro
l
.
2
0
1
8
;
1
6
(6
)
:
2622
-
2
6
2
9
.
[4
]
Ram
a
d
a
n
ZM
.
O
p
ti
m
u
m
I
m
a
g
e
Fi
l
te
r
s
f
o
r
Va
r
i
o
u
s
T
y
p
e
s
o
f
Noi
s
e
.
TEL
KO
M
NIKA
Te
l
e
c
o
m
m
u
n
i
c
a
ti
o
n
Com
p
u
ti
n
g
E
l
e
c
tro
n
i
c
s
a
n
d
Co
n
tro
l
.
2
0
1
8
;
1
6
(
5
)
:
2
4
5
8
-
2464
.
[5
]
Hen
d
ra
w
a
n
Y
,
Sa
k
t
i
IM
,
W
i
b
i
s
o
n
o
Y
,
Rac
h
m
a
w
a
ti
M
,
Su
ta
n
SM
.
Im
a
g
e
A
n
a
l
y
s
i
s
u
s
i
n
g
Co
l
o
r
Co
-
o
c
c
u
rre
n
c
e
M
a
tri
x
T
e
x
tu
ra
l
Fe
a
tu
r
e
s
fo
r
Pr
e
d
i
c
ti
n
g
Nit
r
o
g
e
n
Con
te
n
t
i
n
Sp
i
n
a
c
h
.
TE
L
KO
M
NIKA
Te
l
e
c
o
m
m
u
n
i
c
a
ti
o
n
Co
m
p
u
t
i
n
g
El
e
c
tro
n
i
c
s
a
n
d
Co
n
tr
o
l
.
2
0
1
8
;
1
6
(6
)
:
2
7
1
2
-
2
7
2
4
.
[6
]
Ka
n
g
M
G
,
Cha
u
d
h
u
r
i
S.
Su
p
e
r
-
Res
o
l
u
ti
o
n
I
m
a
g
e
Re
c
o
n
s
tru
c
ti
o
n
.
IEEE
Si
g
n
a
l
Pr
o
c
e
s
s
i
n
g
M
a
g
a
z
i
n
e
.
2
0
0
3
;
2
0
(3
)
:
1
9
–
20.
[7
]
Pa
rk
SC,
Pa
rk
M
K,
Ka
n
g
M
G
.
Su
p
e
r
-
Res
o
l
u
t
i
o
n
I
m
a
g
e
Re
c
o
n
s
tr
u
c
t
i
o
n
:
A
T
e
c
h
n
i
c
a
l
O
v
e
rv
i
e
w
.
IEEE
Si
g
n
a
l
Pro
c
e
s
s
i
n
g
M
a
g
a
z
i
n
e
.
2
0
0
3
;
2
0
(3
)
:
2
1
–
36.
[8
]
S
e
g
a
l
l
CA,
M
o
l
i
n
a
R,
Ka
ts
a
g
g
e
l
o
s
AK.
Hig
h
-
re
s
o
l
u
t
i
o
n
Im
a
g
e
s
fro
m
L
o
w
-
Res
o
l
u
t
i
o
n
C
o
m
p
re
s
s
e
d
Vi
d
e
o
.
IEEE
Si
g
n
a
l
Pr
o
c
e
s
s
i
n
g
M
a
g
a
z
i
n
e
.
2
0
0
3
;
2
0
(3
):
3
7
-
4
8
.
[9
]
Raj
a
n
D,
Cha
u
d
h
u
r
i
S,
J
o
s
h
i
M
V.
M
u
l
ti
-
O
b
j
e
c
t
i
v
e
Su
p
e
r
R
e
s
o
l
u
ti
o
n
Con
c
e
p
t
s
a
n
d
Ex
a
m
p
l
e
s
.
IEE
E
Si
g
n
a
l
Pro
c
e
s
s
i
n
g
M
a
g
a
z
i
n
e
.
2
0
0
3
;
2
0
(3
)
:
49
-
61.
[1
0
]
Ng
M
K,
Bo
s
e
NK.
M
a
th
e
m
a
t
i
c
a
l
a
n
a
l
y
s
i
s
o
f
s
u
p
e
r
-
re
s
o
l
u
ti
o
n
m
e
t
h
o
d
o
l
o
g
y
.
IEEE
Si
g
n
a
l
Pro
c
e
s
s
i
n
g
M
a
g
a
z
i
n
e
.
2
0
0
3
;
2
0
(3
)
:
62
–
74.
[1
1
]
Pa
rk
SC,
P
a
rk
M
K,
Ka
n
g
M
G
.
Su
p
e
r
-
re
s
o
l
u
ti
o
n
i
m
a
g
e
re
c
o
n
s
tru
c
ti
o
n
:
a
te
c
h
n
i
c
a
l
o
v
e
r
v
i
e
w
.
IEEE
Si
g
n
a
l
Pro
c
e
s
s
i
n
g
M
a
g
a
z
i
n
e
.
2
0
0
3
;
2
0
(3
)
:
21
-
36.
[1
2
]
Cha
u
d
h
u
ri
S,
T
a
u
r
DR
.
Hig
h
-
re
s
o
l
u
ti
o
n
s
l
o
w
-
m
o
ti
o
n
s
e
q
u
e
n
c
i
n
g
:
h
o
w
to
g
e
n
e
ra
t
e
a
s
l
o
w
-
m
o
ti
o
n
s
e
q
u
e
n
c
e
f
ro
m
a
b
i
t
s
tre
a
m
.
IE
EE Si
g
n
a
l
Pro
c
e
s
s
i
n
g
M
a
g
a
z
i
n
e
.
2
0
0
5
;
2
2
(
2
):
1
6
-
2
4
.
[1
3
]
Che
n
HM
,
L
e
e
S,
Rao
R
M
,
S
l
a
m
a
n
i
M
A,
Va
rs
h
n
e
y
PK.
Im
a
g
i
n
g
f
o
r
c
o
n
c
e
a
l
e
d
w
e
a
p
o
n
d
e
te
c
ti
o
n
:
a
tu
to
ri
a
l
o
v
e
rv
i
e
w
o
f
d
e
v
e
l
o
p
m
e
n
t
i
n
i
m
a
g
i
n
g
s
e
n
s
o
rs
a
n
d
p
ro
c
e
s
s
i
n
g
.
IEEE
Si
g
n
a
l
Pro
c
e
s
s
i
n
g
M
a
g
a
z
i
n
e
.
2
0
0
5
;
2
2
(5
)
:
1
6
-
24.
[1
4
]
Pa
ta
n
a
v
i
j
i
t
V.
T
u
to
r
i
a
l
o
f
I
m
a
g
e
Rec
o
n
s
tr
u
c
t
i
o
n
B
a
s
e
d
o
n
W
e
i
g
h
t
e
d
Su
m
(
W
S
)
Fi
l
t
e
r
Ap
p
r
o
a
c
h
:
Fro
m
Si
n
g
l
e
I
m
a
g
e
to
M
u
l
ti
-
Fra
m
e
I
m
a
g
e
.
AU J
o
u
rn
a
l
o
f
Te
c
h
n
o
l
o
g
y
(AU J
.T
.)
.
2
0
0
9
;
1
3
(
2
) :
7
5
–
8
6
.
[1
5
]
Pa
ta
n
a
v
i
j
i
t
V
.
Su
p
e
r
-
R
e
s
o
l
u
t
i
o
n
Rec
o
n
s
tr
u
c
ti
o
n
a
n
d
i
ts
F
u
tu
re
Res
e
a
r
c
h
Dir
e
c
ti
o
n
.
A
U
J
o
u
rn
a
l
o
f
Te
c
h
n
o
l
o
g
y
(AU
J
.T
.)
.
2
0
0
9
;
1
2
(3
):
1
4
9
–
1
6
3
.
[1
6
]
Pa
ta
n
a
v
i
j
i
t
V.
M
a
th
e
m
a
ti
c
a
l
An
a
l
y
s
i
s
o
f
St
o
c
h
a
s
t
i
c
R
e
g
u
l
a
ri
z
a
ti
o
n
Ap
p
ro
a
c
h
fo
r
Su
p
e
r
-
Res
o
l
u
ti
o
n
Rec
o
n
s
tru
c
ti
o
n
.
AU
J
o
u
rn
a
l
o
f
Te
c
h
n
o
l
o
g
y
(AU
J
.T
.)
.
2
0
0
9
;
1
2
(4
)
:
2
3
5
–
2
4
4
.
[1
7
]
Pa
ta
n
a
v
i
j
i
t
V.
Co
m
p
u
ta
ti
o
n
a
l
T
u
to
ri
a
l
o
f
S
te
e
p
e
s
t
De
s
c
e
n
t
M
e
th
o
d
a
n
d
It
s
I
m
p
l
e
m
e
n
ta
ti
o
n
i
n
Di
g
i
ta
l
Im
a
g
e
Pro
c
e
s
s
i
n
g
.
ECTI
E
-
m
a
g
a
z
i
n
e
.
2013
;
7
(1
)
.
[1
8
]
J
e
o
n
BW
,
e
t
a
l
.
Res
o
l
u
t
i
o
n
E
n
h
a
n
c
e
m
e
n
t
b
y
Pre
d
i
c
ti
o
n
o
f
th
e
Hig
h
-
Fr
e
q
u
e
n
c
y
Im
a
g
e
B
a
s
e
d
o
n
th
e
L
a
p
l
a
c
i
a
n
Py
ra
m
i
d
.
EUR
ASIP
J
o
u
rn
a
l
o
n
Ad
v
a
n
c
e
s
i
n
S
i
g
n
a
l
Pro
c
e
s
s
i
n
g
(
J
ASP).
2
0
0
6
:
1
-
11
.
[1
9
]
Bu
rt
P,
A
d
e
l
s
o
n
E.
T
h
e
L
a
p
l
a
c
i
a
n
p
y
ra
m
i
d
a
s
a
c
o
m
p
a
c
t
i
m
a
g
e
c
o
d
e
.
IEEE
Tra
n
s
a
c
t
i
o
n
s
o
n
Com
m
u
n
i
c
a
ti
o
n
.
1
9
8
3
;
3
1
(
4
):
532
–
5
4
0
.
[2
0
]
L
i
X
,
O
rc
h
a
rd
M
T
.
New
e
d
g
e
-
d
i
re
c
te
d
i
n
t
e
rp
o
l
a
ti
o
n
.
IEEE
T
ra
n
s
a
c
ti
o
n
s
o
n
Im
a
g
e
Pr
o
c
e
s
s
i
n
g
.
2
0
0
1
;
1
0
(1
0
)
:
1
5
2
1
–
1
5
2
7
.
[2
1
]
Bi
a
n
c
a
rd
i
A,
Cin
q
u
e
L
,
L
o
m
b
a
rd
i
L
.
Im
p
ro
v
e
m
e
n
ts
to
i
m
a
g
e
m
a
g
n
i
fi
c
a
ti
o
n
.
P
a
tt
e
r
n
Rec
o
g
n
i
ti
o
n
.
2
0
0
2
;
3
5
(3
):
677
–
687.
[2
2
]
L
e
u
J
G
.
Sh
a
rp
n
e
s
s
p
re
s
e
rv
i
n
g
i
m
a
g
e
e
n
l
a
rg
e
m
e
n
t
b
a
s
e
d
o
n
a
ra
m
p
e
d
g
e
m
o
d
e
l
.
Pa
tt
e
r
n
R
e
c
o
g
n
i
ti
o
n
.
2
0
0
1
;
3
4
(1
0
)
:
1
9
2
7
–
1
9
3
8
.
[2
3
]
W
a
n
g
Q
,
W
a
rd
R.
A
c
o
n
to
u
r
-
p
re
s
e
r
v
i
n
g
i
m
a
g
e
i
n
te
rp
o
l
a
ti
o
n
m
e
th
o
d
.
Pro
c
e
e
d
i
n
g
s
2
0
0
3
In
te
rn
a
t
i
o
n
a
l
Con
fe
re
n
c
e
o
n
I
m
a
g
e
Pro
c
e
s
s
i
n
g
(I
CIP
’
0
3
).
Ba
rc
e
l
o
n
a
.
2
0
0
3
;
3
:
6
7
3
–
676.
[2
4
]
Pa
ta
n
a
v
i
j
i
t
V
,
P
i
ra
k
C,
As
c
h
e
i
d
G.
E
x
p
e
ri
m
e
n
ta
l
Pe
rf
o
rm
a
n
c
e
An
a
l
y
s
i
s
o
f
Im
a
g
e
Hi
g
h
-
fre
q
u
e
n
c
y
i
m
a
g
e
p
re
d
i
c
ti
o
n
re
c
o
n
s
tru
c
ti
o
n
B
a
s
e
d
o
n
th
e
Hi
g
h
-
Fr
e
q
u
e
n
c
y
Im
a
g
e
Pre
d
i
c
t
i
o
n
u
n
d
e
r
Se
v
e
ra
l
Bl
u
rre
d
a
n
d
Noi
s
y
En
v
i
ro
n
m
e
n
ts
.
T
h
e
I
EEE
In
te
r
n
a
ti
o
n
a
l
Sy
m
p
o
s
i
u
m
o
n
Co
m
m
u
n
i
c
a
ti
o
n
s
a
n
d
In
fo
r
m
a
t
i
o
n
T
e
c
h
n
o
l
o
g
i
e
s
(I
SCIT
2
0
1
3
)
.
T
h
a
i
l
a
n
d
.
2
0
1
3
.
[2
5
]
Pa
ta
n
a
v
i
j
i
t
V,
Pi
ra
k
C,
A
s
c
h
e
i
d
G
.
A
Pe
rf
o
rm
a
n
c
e
Im
p
a
c
t
o
f
An
Ed
g
e
Ke
r
n
e
l
fo
r
T
h
e
Hi
g
h
-
Fre
q
u
e
n
c
y
Im
a
g
e
Pre
d
i
c
ti
o
n
Rec
o
n
s
t
ru
c
ti
o
n
.
T
h
e
IEEE
In
te
rn
a
ti
o
n
a
l
Sy
m
p
o
s
i
u
m
o
n
Com
m
u
n
i
c
a
ti
o
n
s
a
n
d
In
fo
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
i
e
s
(ISC
IT
2
0
1
4
)
.
In
c
h
e
o
n
.
2
0
1
4
.
[2
6
]
Bl
a
c
k
M
J
,
Ran
g
a
ra
j
a
n
A.
O
n
th
e
u
n
i
fi
c
a
ti
o
n
o
f
l
i
n
e
p
r
o
c
e
s
s
e
s
,
o
u
tl
i
e
r
re
j
e
c
t
i
o
n
,
a
n
d
ro
b
u
s
t
s
t
a
ti
s
ti
c
s
w
i
th
a
p
p
l
i
c
a
t
i
o
n
s
i
n
e
a
r
l
y
v
i
s
i
o
n
.
In
te
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Co
m
p
u
te
r Vi
s
i
o
n
.
1
9
9
6
;
1
9
(1
)
:
57
-
91
.
Evaluation Warning : The document was created with Spire.PDF for Python.