I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
7
,
No
.
4
,
A
u
g
u
s
t
201
7
,
p
p
.
1
9
6
4
~
1
9
7
2
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v7
i
4
.
p
p
1
9
6
4
-
1972
1964
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
jo
u
r
n
a
l.c
o
m/o
n
lin
e/in
d
ex
.
p
h
p
/I
JE
C
E
Dictiona
ry
bas
ed
I
m
a
g
e Co
m
press
i
o
n via
Spa
rse Re
presenta
tion
Ara
bin
da
Sa
ho
o
1
,
P
ra
na
t
i D
a
s
2
1
De
p
a
rtme
n
t
o
f
ECE
,
IT
ER,
S
ik
sh
a
„O‟ A
n
u
sa
n
d
h
a
n
Un
iv
e
rsity
,
Bh
u
b
a
n
e
sw
a
r,
Od
ish
a
,
In
d
ia
2
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
IG
I
T
,
S
a
ra
n
g
,
Od
ish
a
,
I
n
d
i
a
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
2
7
,
2
0
1
7
R
ev
i
s
ed
Ma
y
2
5
,
2
0
1
7
A
cc
ep
ted
J
u
n
1
5
,
2
0
1
7
No
w
a
d
a
y
s
ima
g
e
c
o
m
p
re
ss
io
n
h
a
s
b
e
c
o
m
e
a
n
e
c
e
ss
it
y
d
u
e
to
a
lar
g
e
v
o
lu
m
e
o
f
i
m
a
g
e
s.
F
o
r
e
ff
icie
n
t
u
se
o
f
sto
ra
g
e
sp
a
c
e
a
n
d
d
a
ta
tran
s
m
is
sio
n
,
it
b
e
c
o
m
e
s
e
ss
e
n
ti
a
l
to
c
o
m
p
re
ss
th
e
im
a
g
e
.
In
th
is
p
a
p
e
r,
w
e
p
ro
p
o
se
a
d
ictio
n
a
ry
b
a
se
d
ima
g
e
c
o
m
p
re
s
sio
n
f
ra
m
e
w
o
r
k
v
ia
sp
a
rse
re
p
r
e
se
n
tatio
n
,
w
it
h
th
e
c
o
n
stru
c
ti
o
n
o
f
a
trai
n
e
d
o
v
e
r
-
c
o
m
p
lete
d
ictio
n
a
ry
.
T
h
e
o
v
e
r
-
c
o
m
p
lete
d
ictio
n
a
ry
is
tr
a
in
e
d
u
sin
g
th
e
in
tra
-
p
re
d
ictio
n
re
sid
u
a
l
s
o
b
tain
e
d
f
ro
m
d
iff
e
re
n
t
i
m
a
g
e
s
a
n
d
is
a
p
p
li
e
d
f
o
r
sp
a
rse
re
p
re
se
n
tatio
n
.
I
n
t
h
is
m
e
th
o
d
,
t
h
e
c
u
rre
n
t
im
a
g
e
b
l
o
c
k
is
f
irst
p
re
d
icte
d
f
ro
m
it
s
sp
a
ti
a
ll
y
n
e
ig
h
b
o
r
in
g
b
lo
c
k
s,
a
n
d
th
e
n
th
e
p
re
d
ictio
n
re
sid
u
a
ls
a
re
e
n
c
o
d
e
d
v
ia
sp
a
rs
e
re
p
re
se
n
tatio
n
.
S
p
a
rse
a
p
p
ro
x
i
m
a
ti
o
n
a
lg
o
rit
h
m
a
n
d
th
e
tra
in
e
d
o
v
e
r
-
c
o
m
p
lete
d
ictio
n
a
ry
a
re
a
p
p
li
e
d
f
o
r
sp
a
rse
re
p
re
s
e
n
tatio
n
o
f
p
re
d
icti
o
n
re
sid
u
a
ls.
T
h
e
d
e
tail
c
o
e
ff
icie
n
t
s
o
b
tain
e
d
f
ro
m
sp
a
rse
r
e
p
re
se
n
tatio
n
a
re
u
se
d
f
o
r
e
n
c
o
d
i
n
g
.
Ex
p
e
rim
e
n
ta
l
re
su
lt
sh
o
w
s
th
a
t
th
e
p
r
o
p
o
se
d
m
e
th
o
d
y
i
e
ld
s
b
o
th
im
p
ro
v
e
d
c
o
d
i
n
g
e
fficie
n
c
y
a
n
d
ima
g
e
q
u
a
li
ty
a
s
c
o
m
p
a
re
d
to
so
m
e
s
tate
-
of
-
th
e
-
ar
t
i
m
a
g
e
co
m
p
r
e
s
s
io
n
m
et
h
o
d
s
.
K
ey
w
o
r
d
:
Dictio
n
ar
y
lear
n
i
n
g
I
m
ag
e
co
m
p
r
es
s
io
n
I
n
tr
a
p
r
ed
ictio
n
K
-
SV
D
Sp
ar
s
e
r
ep
r
esen
tatio
n
Co
p
y
rig
h
t
©
2
0
1
7
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
A
r
ab
in
d
a
Sa
h
o
o
,
Dep
ar
te
m
en
t o
f
E
lectr
o
n
ics a
n
d
C
o
m
m
u
n
icatio
n
E
n
g
in
ee
r
i
n
g
,
I
T
E
R
,
Sik
s
h
a
„
O
‟
A
n
u
s
a
n
d
h
a
n
Un
i
v
er
s
it
y
,
B
h
u
b
an
e
s
w
ar
,
Od
is
h
a,
I
n
d
ia.
E
m
ail: a
r
u
b
ab
u
1
2
3
@
y
a
h
o
o
.
co
.
in
1.
I
NT
RO
D
UCT
I
O
N
I
m
ag
e
s
co
m
p
r
ess
io
n
[
1
]
h
as
a
l
w
a
y
s
b
ee
n
i
m
p
o
r
tan
t,
an
d
n
o
w
ad
a
y
s
it
b
ec
o
m
e
s
es
s
en
tial
d
u
e
to
th
e
h
u
g
e
r
eq
u
ir
e
m
e
n
t
o
f
i
m
a
g
e
s
to
r
ag
e
an
d
tr
an
s
f
er
.
O
v
er
t
h
e
last
t
w
o
d
ec
ad
es,
n
u
m
er
o
u
s
an
d
d
iv
er
s
e
i
m
ag
e
co
m
p
r
es
s
io
n
m
et
h
o
d
s
[2
-
6]
h
av
e
b
ee
n
p
r
o
p
o
s
ed
.
T
h
e
m
o
s
t
w
id
el
y
u
s
ed
m
e
th
o
d
s
ar
e
b
ased
o
n
tr
an
s
f
o
r
m
-
b
ased
co
d
in
g
.
B
ased
o
n
tr
an
s
f
o
r
m
-
ba
s
ed
ap
p
r
o
ac
h
es
b
y
f
ar
m
an
y
i
m
a
g
e
co
m
p
r
es
s
io
n
s
tan
d
ar
d
s
lik
e
J
P
E
G
[
2
]
an
d
J
P
E
G2
0
0
0
[
3
]
h
av
e
b
ee
n
d
ev
elo
p
ed
.
T
h
e
w
id
el
y
u
s
ed
J
P
E
G
an
d
J
P
E
G2
0
0
0
s
tan
d
ar
d
s
u
tili
ze
d
is
cr
ete
co
s
in
e
tr
a
n
s
f
o
r
m
(
DC
T
)
an
d
w
av
ele
t
tr
a
n
s
f
o
r
m
to
co
m
p
r
ess
i
v
el
y
r
ep
r
ese
n
t
i
m
a
g
es.
H
o
w
e
v
er
,
J
P
E
G
an
d
J
P
E
G2
0
0
0
d
o
n
o
t
tak
e
s
p
atial
co
r
r
elatio
n
o
f
n
eig
h
b
o
r
in
g
b
lo
ck
s
in
to
co
n
s
id
er
atio
n
f
o
r
m
o
r
e
co
m
p
ac
t
r
ep
r
esen
tatio
n
o
f
i
m
a
g
e
s
[
7
]
.
I
n
r
ec
en
t
y
ea
r
s
,
in
tr
a
-
p
r
ed
ictio
n
[
8
]
,
[
9
]
s
ch
em
e
s
s
er
v
e
as
a
p
r
o
m
i
s
i
n
g
d
ir
ec
tio
n
,
w
h
ic
h
u
s
e
n
e
ig
h
b
o
r
in
g
b
lo
ck
s
in
t
h
e
co
m
p
r
ess
io
n
p
r
o
ce
s
s
f
o
r
m
o
r
e
co
m
p
r
ess
io
n
.
T
h
e
id
ea
o
f
in
tr
a
-
p
r
ed
ictio
n
is
to
p
r
ed
ict
th
e
u
n
k
n
o
w
n
i
m
ag
e
b
lo
ck
b
ased
o
n
th
e
k
n
o
w
led
g
e
o
f
d
ec
o
d
ed
n
eig
h
b
o
r
in
g
b
lo
ck
s
.
A
g
o
o
d
p
r
ed
ictio
n
r
ed
u
ce
s
th
e
o
v
er
all
co
d
in
g
r
ate.
Ho
w
e
v
er
,
th
ese
co
m
p
r
ess
io
n
m
et
h
o
d
s
s
u
f
f
er
s
o
m
e
li
m
it
atio
n
s
,
b
ec
au
s
e
th
e
y
ar
e
n
o
t
ab
le
to
ef
f
icien
tl
y
co
m
p
r
ess
s
o
m
e
s
p
ec
if
ic
cla
s
s
es
o
f
i
m
a
g
es.
T
h
e
y
ar
e
n
o
t
ab
le
to
s
p
ar
s
el
y
r
ep
r
esen
t
co
m
p
le
x
ch
ar
ac
ter
is
tic
s
o
f
a
n
i
m
a
g
e.
T
o
o
v
er
co
m
e
th
e
s
e
li
m
ita
t
io
n
s
,
s
p
ar
s
e
r
ep
r
esen
tatio
n
s
[
1
0
]
h
as
b
ee
n
e
v
o
lv
i
n
g
i
n
r
ec
en
t
y
ea
r
s
.
Sp
ar
s
e
r
ep
r
ese
n
t
atio
n
i
s
a
v
er
y
e
f
f
ec
ti
v
e
to
o
l
f
o
r
co
m
p
r
es
s
in
g
a
lar
g
e
v
ar
iet
y
o
f
i
m
a
g
es.
T
h
is
i
s
p
o
s
s
ib
le
d
u
e
to
t
h
e
f
ac
t t
h
at
th
e
i
m
a
g
es
ca
n
b
e
s
p
ar
s
e
o
r
co
m
p
r
ess
ib
le
w
it
h
r
esp
ec
t
to
s
o
m
e
b
asis
o
r
d
ictio
n
ar
y
[
1
1
]
.
T
h
u
s
,
s
p
ar
s
e
r
ep
r
esen
t
atio
n
p
r
o
v
id
es
a
p
o
ten
tial
f
o
r
ef
f
ec
ti
v
e
i
m
ag
e
co
m
p
r
ess
io
n
.
An
i
m
a
g
e
is
co
m
p
r
es
s
ib
le
o
r
n
o
t
,
it d
ep
en
d
s
o
n
th
e
d
ictio
n
ar
y
,
s
o
th
e
d
esi
g
n
o
f
d
ictio
n
ar
y
i
s
v
i
tal
in
s
p
ar
s
e
r
ep
r
esen
tatio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Dictio
n
a
r
y
b
a
s
ed
I
ma
g
e
C
o
mp
r
ess
io
n
via
S
p
a
r
s
e
R
ep
r
esen
ta
tio
n
(
A
r
a
b
in
d
a
S
a
h
o
o
)
1965
I
n
co
n
tr
ast
to
f
i
x
ed
D
C
T
an
d
w
a
v
elet
d
ictio
n
ar
y
,
t
h
e
late
s
t
t
r
en
d
o
f
i
m
ag
e
co
m
p
r
ess
io
n
te
ch
n
iq
u
es
i
s
ex
ten
d
ed
to
d
esig
n
tr
ai
n
ed
d
ictio
n
ar
ies
[
1
2
]
.
Nu
m
er
o
u
s
d
ic
tio
n
ar
y
b
ased
i
m
ag
e
co
m
p
r
e
s
s
io
n
m
et
h
o
d
s
h
a
v
e
b
ee
n
p
r
o
p
o
s
ed
f
o
r
s
p
ar
s
e
r
e
p
r
esen
tatio
n
.
I
n
s
e
v
er
al
r
ec
en
tl
y
p
u
b
lis
h
ed
w
o
r
k
s
,
th
e
u
s
e
o
f
lear
n
ed
o
v
e
r
-
co
m
p
lete
d
ictio
n
ar
ies
in
i
m
a
g
e
co
m
p
r
ess
io
n
h
a
s
s
h
o
w
n
p
r
o
m
is
in
g
r
esu
lts
at
lo
w
b
it
r
ate
as
co
m
p
ar
ed
to
f
ix
ed
d
ictio
n
ar
ies.
T
h
e
in
itial
p
r
o
p
o
s
al
to
w
ar
d
s
d
ictio
n
ar
y
b
ase
d
im
a
g
e
co
m
p
r
ess
io
n
w
as
p
r
o
p
o
s
ed
b
y
B
r
y
t
an
d
E
lad
[
1
3
]
.
T
h
ey
p
r
o
p
o
s
ed
an
a
lg
o
r
ith
m
f
o
r
t
h
e
co
m
p
r
es
s
io
n
o
f
f
ac
ial
i
m
a
g
e
b
ased
o
n
th
e
l
ea
r
n
ed
K
-
SVD
[
1
4
]
o
v
er
-
co
m
p
lete
d
ictio
n
ar
y
.
T
h
o
u
g
h
t
h
is
m
et
h
o
d
o
u
tp
er
f
o
r
m
s
J
P
E
G
an
d
J
P
E
G2
0
0
0
b
u
t
th
is
m
et
h
o
d
is
li
m
ited
to
c
o
m
p
r
es
s
in
g
f
ac
ial
i
m
a
g
es
.
I
n
[
1
5
]
,
th
e
au
t
h
o
r
p
r
o
p
o
s
ed
an
i
m
a
g
e
co
m
p
r
e
s
s
io
n
m
et
h
o
d
b
ase
d
o
n
th
e
iter
atio
n
-
tu
n
ed
d
ictio
n
ar
y
(
I
T
D)
.
I
n
th
i
s
s
c
h
e
m
e,
t
h
e
d
ictio
n
ar
y
co
n
s
i
s
ts
o
f
a
la
y
er
s
tr
u
ct
u
r
e
w
it
h
ea
c
h
la
y
er
tr
ain
ed
f
o
r
a
s
p
ec
if
ic
class
o
f
i
m
a
g
es
a
n
d
ca
r
r
ies
a
s
ep
ar
ate
d
ictio
n
ar
y
m
atr
i
x
.
T
h
is
m
et
h
o
d
is
s
h
o
w
n
to
o
u
tp
er
f
o
r
m
K
-
SV
D
o
v
er
-
co
m
p
lete
d
ictio
n
ar
y
m
et
h
o
d
b
u
t
it
is
e
m
p
lo
y
ed
to
co
m
p
r
ess
a
s
p
ec
if
ic
cla
s
s
o
f
i
m
a
g
es.
I
n
[
1
6
]
,
th
e
a
u
th
o
r
p
r
o
p
o
s
ed
an
ad
ap
tiv
e
d
icti
o
n
ar
y
d
esi
g
n
m
et
h
o
d
f
o
r
th
e
f
i
n
g
er
p
r
in
t
i
m
a
g
e
co
m
p
r
es
s
io
n
.
T
h
e
y
u
s
ed
a
s
et
o
f
f
i
n
g
er
p
r
in
t
i
m
a
g
e
to
lear
n
a
d
ictio
n
ar
y
.
I
n
[
1
7
]
,
th
e
a
u
th
o
r
p
r
o
p
o
s
ed
a
d
ictio
n
ar
y
b
ased
m
et
h
o
d
f
o
r
co
m
p
r
es
s
i
n
g
s
u
r
v
eilla
n
ce
i
m
ag
e.
E
x
p
er
i
m
e
n
tal
r
e
s
u
l
t
s
h
o
w
s
t
h
ese
ab
o
v
e
p
r
o
p
o
s
ed
alg
o
r
it
h
m
s
o
u
tp
er
f
o
r
m
J
P
E
G
an
d
J
P
E
G2
0
0
0
.
Ho
w
e
v
er
,
m
o
s
t
o
f
t
h
e
ab
o
v
e
m
en
tio
n
ed
co
m
p
r
es
s
io
n
s
ch
e
m
e
is
e
ith
er
r
elate
d
to
f
ac
ial
i
m
a
g
e
s
o
r
to
s
o
m
e
s
p
ec
if
ic
cla
s
s
o
f
i
m
ag
e
s
,
w
h
il
e
th
er
e
is
a
lack
o
f
r
esea
r
ch
o
n
g
e
n
er
al
ar
b
itra
r
y
i
m
a
g
es
.
So
,
th
e
m
ain
ch
alle
n
g
e
o
f
ab
o
v
e
s
c
h
e
m
es i
s
t
h
e
co
m
p
r
ess
io
n
o
f
g
e
n
er
al
ar
b
itra
r
y
i
m
ag
es.
T
o
ad
d
r
ess
th
is
,
i
n
t
h
i
s
p
ap
er
w
e
p
r
o
p
o
s
ed
a
n
o
v
el
i
m
a
g
e
co
m
p
r
e
s
s
i
o
n
s
c
h
e
m
e
f
o
r
ar
b
itra
r
y
i
m
a
g
e
s
.
I
n
th
is
s
c
h
e
m
e,
an
e
f
f
icie
n
t
tr
ai
n
ed
o
v
er
-
co
m
p
lete
d
ictio
n
ar
y
is
i
n
te
g
r
ated
in
to
th
e
in
tr
a
-
p
r
ed
ictio
n
f
r
a
m
e
w
o
r
k
.
T
h
e
co
n
v
e
n
tio
n
a
l
tr
an
s
f
o
r
m
-
d
o
m
ai
n
r
ep
r
esen
tatio
n
o
f
i
n
tr
a
-
p
r
ed
ictio
n
s
c
h
e
m
e
i
s
r
ep
lace
d
b
y
a
tr
ain
ed
o
v
er
-
co
m
p
l
ete
d
ictio
n
ar
y
.
W
e
tr
ai
n
ed
a
d
ictio
n
ar
y
o
f
f
li
n
e
u
s
i
n
g
th
e
r
esid
u
als
o
b
tain
ed
f
r
o
m
i
n
tr
a
-
p
r
ed
ictio
n
.
T
h
is
d
ictio
n
ar
y
ca
n
s
p
ar
s
el
y
r
ep
r
esen
t
th
e
co
m
p
le
x
ch
ar
ac
ter
i
s
tics
o
f
t
h
e
r
esid
u
al
b
lo
c
k
.
T
h
e
co
ef
f
icien
ts
a
n
d
in
d
ices
o
f
ap
p
r
o
p
r
iate
d
ictio
n
ar
y
ele
m
e
n
t
o
b
tain
ed
f
r
o
m
s
p
ar
s
e
r
ep
r
esen
tatio
n
ar
e
tr
an
s
m
itte
d
f
o
r
en
co
d
in
g
.
Sin
ce
th
e
d
ic
tio
n
ar
y
is
s
h
ar
ed
at
b
o
th
en
co
d
er
an
d
d
ec
o
d
er
,
o
n
l
y
co
ef
f
icie
n
ts
a
n
d
in
d
ices
o
f
d
ictio
n
ar
y
ele
m
e
n
ts
n
ee
d
to
b
e
en
co
d
ed
,
w
h
ich
co
m
p
r
es
s
th
e
i
m
ag
e
s
ig
n
i
f
ica
n
tl
y
.
E
x
p
er
i
m
e
n
tal
r
esu
lt
s
on
th
e
ar
b
itra
r
y
i
m
ag
e
s
s
h
o
w
s
th
at
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
y
ield
s
b
o
th
i
m
p
r
o
v
ed
co
d
in
g
ef
f
icie
n
c
y
an
d
i
m
ag
e
q
u
alit
y
as c
o
m
p
ar
ed
to
J
P
E
G
a
n
d
J
P
E
G
2
0
0
0
.
T
h
e
r
est
o
f
th
e
p
ap
er
is
o
r
g
an
ized
as
f
o
llo
w
s
.
I
n
Sectio
n
2
,
w
e
p
r
esen
t
s
o
m
e
p
r
eli
m
in
ar
ie
s
o
n
s
p
ar
s
e
r
ep
r
e
s
en
tatio
n
a
n
d
d
ictio
n
ar
y
d
esig
n
.
T
h
e
p
r
o
p
o
s
ed
i
m
a
g
e
co
m
p
r
ess
io
n
m
et
h
o
d
is
i
n
tr
o
d
u
ce
d
in
s
ec
tio
n
3
.
Sectio
n
4
illu
s
tr
ates e
x
p
er
i
m
e
n
tal
r
es
u
lts
a
n
d
d
is
cu
s
s
io
n
.
Fi
n
all
y
,
Sectio
n
5
co
n
cl
u
d
es t
h
e
p
ap
er
.
2.
P
RE
L
I
M
I
NARIE
S
I
n
s
p
ar
s
e
r
ep
r
esen
tatio
n
[
1
8
]
,
a
s
ig
n
al
ca
n
b
e
r
ep
r
esen
ted
b
y
a
li
n
ea
r
co
m
b
in
at
io
n
o
f
a
s
m
all
n
u
m
b
er
o
f
s
i
g
n
a
ls
k
n
o
w
n
a
s
ato
m
s
ta
k
en
f
r
o
m
a
n
o
v
er
-
co
m
p
lete
d
ictio
n
ar
y
.
I
t
is
ca
lled
s
p
ar
s
e
r
ep
r
esen
tatio
n
as
it
e
m
p
lo
y
s
o
n
l
y
a
f
e
w
n
u
m
b
er
d
ictio
n
ar
y
a
to
m
s
o
r
ele
m
e
n
t
s
to
r
ep
r
esen
t
th
e
s
i
g
n
al.
A
s
i
g
n
al
is
s
aid
to
b
e
co
m
p
r
ess
ib
le
if
it
ca
n
b
e
r
ep
r
esen
ted
b
y
f
e
w
d
ict
io
n
ar
y
ato
m
s
.
Ma
t
h
e
m
a
ticall
y
s
p
ar
s
e
r
ep
r
esen
tatio
n
ca
n
b
e
ex
p
r
ess
e
d
as
:
(
1
)
T
h
e
s
o
lu
tio
n
v
ec
to
r
co
n
ta
in
s
th
e
co
e
f
f
icie
n
ts
o
f
t
h
e
s
i
g
n
al
.
W
h
er
e
D
i
s
o
n
e
n
×
k
m
atr
i
x
w
it
h
n
<
k
ca
lled
o
v
er
-
co
m
p
lete
d
ictio
n
ar
y
a
n
d
ea
ch
co
lu
m
n
o
f
D
is
ca
lled
an
ato
m
.
I
f
an
d
ea
ch
ato
m
o
f
D
ar
e
tr
ea
te
d
as
a
s
i
g
n
al
t
h
e
n
ca
n
b
e
r
ep
r
esen
ted
as
a
li
n
ea
r
co
m
b
in
a
tio
n
o
f
ato
m
s
o
f
D
.
T
h
is
li
n
ea
r
co
m
b
in
at
io
n
ca
n
b
e
e
x
p
r
ess
ed
as
s
o
lu
tio
n
v
ec
to
r
.
Du
e
to
o
v
er
co
m
p
lete
n
at
u
r
e
o
f
D
,
an
in
f
in
i
te
n
u
m
b
er
o
f
s
o
lu
tio
n
e
x
is
t
f
o
r
.
I
n
last
d
ec
ad
e
,
v
ar
io
u
s
s
p
ar
s
e
ap
p
r
o
x
i
m
atio
n
alg
o
r
it
h
m
s
h
a
v
e
b
ee
n
p
r
o
p
o
s
ed
to
f
in
d
o
u
t
th
e
s
p
ar
s
e
s
o
l
u
tio
n
f
o
r
.
T
h
e
s
p
ar
s
e
ap
p
r
o
x
i
m
atio
n
al
g
o
r
ith
m
al
w
a
y
s
ai
m
s
to
r
e
p
r
esen
t
in
ter
m
s
o
f
m
in
i
m
u
m
n
u
m
b
er
ato
m
s
.
Ma
th
e
m
atica
ll
y
,
t
h
is
ca
n
b
e
ex
p
r
ess
ed
as
s
o
l
v
i
n
g
E
q
u
a
tio
n
(
1
)
s
u
ch
t
h
at
th
e
s
o
lu
t
io
n
co
n
tain
s
m
in
i
m
u
m
n
u
m
b
er
o
f
n
o
n
-
ze
r
o
ele
m
e
n
ts
.
A
s
ig
n
al
i
s
s
aid
to
b
e
co
m
p
r
es
s
ib
le
i
f
th
e
n
u
m
b
er
o
f
n
o
n
-
ze
r
o
elem
e
n
t
s
in
is
v
er
y
le
s
s
a
s
co
m
p
ar
ed
to
n
u
m
b
er
o
f
ele
m
en
ts
o
f
.
T
h
e
s
p
ar
s
e
r
ep
r
esen
tatio
n
o
f
m
a
y
b
e
eith
er
ex
ac
t
or
ap
p
r
o
x
i
m
ate,
s
atis
f
y
i
n
g
.
W
h
er
e
p
is
th
e
n
o
r
m
.
T
y
p
ical
n
o
r
m
s
u
s
ed
i
n
t
h
e
ap
p
r
o
x
im
a
tio
n
ar
e
1
,
2
o
r
∞.
No
r
m
a
ll
y
p
=
2
is
ta
k
en
in
i
m
ag
e
co
m
p
r
es
s
io
n
.
Ma
th
e
m
at
icall
y
,
t
h
e
s
p
r
ar
s
est
r
ep
r
esen
tatio
n
is
t
h
e
s
o
l
u
tio
n
o
f
eit
h
er
E
q
u
atio
n
s
(
2
)
o
r
(
3
)
(
2
)
o
r
(
3
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
4
,
A
u
g
u
s
t
2017
:
1
9
6
4
–
1
9
7
2
1966
w
h
er
e
is
th
e
l
0
n
o
r
m
,
r
ep
r
esen
ts
th
e
n
u
m
b
er
o
f
n
o
n
-
ze
r
o
elem
en
ts
in
s
o
l
u
tio
n
v
ec
to
r
x
.
So
m
e
o
f
th
e
w
el
l
k
n
o
w
n
al
g
o
r
ith
m
s
u
s
ed
to
f
i
n
d
s
p
ar
s
e
r
ep
r
esen
tatio
n
ar
e:
Ma
tch
i
n
g
p
u
r
s
u
it
(
MP
)
[
1
9
]
,
Or
th
o
g
o
n
a
l
Ma
tch
i
n
g
P
u
r
s
u
it
(
OM
P
)
[
2
0
]
an
d
C
o
m
p
li
m
en
tar
y
m
a
tch
i
n
g
p
u
r
s
u
i
t
(
C
MP
)
[
2
1
]
.
I
n
th
is
p
ap
er
,
w
e
f
o
cu
s
o
n
O
MP
alg
o
r
ith
m
d
u
e
to
its
e
f
f
icie
n
c
y
.
T
h
e
d
ictio
n
ar
y
b
ased
i
m
a
g
e
c
o
m
p
r
es
s
io
n
ca
n
also
b
e
ef
f
ec
t
iv
el
y
m
o
d
eled
b
y
E
q
u
a
tio
n
s
(
2
)
an
d
(
3
)
.
I
n
i
m
ag
e
co
m
p
r
ess
io
n
,
w
e
co
n
s
id
er
a
s
et
o
f
k
i
m
ag
e
b
lo
ck
s
o
f
s
ize
n
p
ix
el
s
w
i
t
h
n
<
k
,
o
r
d
er
ed
lex
ico
g
r
ap
h
icall
y
as
co
lu
m
n
v
ec
to
r
s
o
f
d
ictio
n
ar
y
D
.
A
co
lu
m
n
v
ec
to
r
y
is
o
b
tain
ed
f
r
o
m
an
i
m
a
g
e
b
lo
ck
o
f
s
ize
n
p
ix
els.
I
n
s
p
ar
s
e
r
ep
r
esen
tatio
n
,
t
h
e
p
r
o
b
lem
is
to
f
i
n
d
o
u
t
th
e
s
o
lu
tio
n
v
ec
to
r
w
h
i
ch
w
ill
r
ep
r
esen
t
y
w
it
h
lea
s
t
n
u
m
b
er
o
f
d
ictio
n
ar
y
ele
m
e
n
ts
.
I
n
d
ee
d
,
co
m
p
r
ess
io
n
o
f
i
m
a
g
e
p
atch
y
ca
n
b
e
ac
h
ie
v
ed
b
y
tr
an
s
m
is
s
io
n
o
f
n
o
n
ze
r
o
ele
m
e
n
ts
o
f
v
ec
to
r
,
b
y
s
p
ec
if
y
i
n
g
t
h
eir
co
ef
f
icie
n
ts
a
n
d
in
d
ices
.
T
h
e
d
ictio
n
ar
y
p
la
y
s
a
n
i
m
p
o
r
tan
t
r
o
le
in
a
s
u
cc
e
s
s
f
u
l
i
m
ag
e
co
m
p
r
e
s
s
io
n
m
o
d
eli
n
g
v
ia
s
p
ar
s
e
r
ep
r
esen
tatio
n
.
Am
i
m
a
g
e
is
c
o
m
p
r
es
s
ib
le
i
f
it
ca
n
b
e
r
ep
r
esen
ted
b
y
f
e
w
n
u
m
b
er
o
f
d
ictio
n
ar
y
ele
m
e
n
ts
.
T
h
e
d
ictio
n
ar
y
ca
n
eit
h
er
b
e
ch
o
s
e
n
as
a
p
r
esp
ec
if
ied
s
et
o
f
i
m
a
g
es
o
r
d
esig
n
ed
b
y
ad
ap
tin
g
i
ts
co
n
te
n
ts
to
f
it
a
g
iv
e
n
s
et
o
f
i
m
a
g
es.
T
h
e
o
b
jectiv
e
o
f
d
ictio
n
ar
y
d
e
s
ig
n
is
to
tr
ain
th
e
d
ictio
n
ar
y
w
h
ic
h
ab
le
to
r
e
p
r
esen
t
a
s
ig
n
al
s
e
t
s
p
ar
s
el
y
[
1
2
]
.
Giv
e
n
an
i
m
a
g
e
s
et
{
}
,
d
ictio
n
ar
y
d
esi
g
n
ai
m
s
to
f
i
n
d
th
e
b
est
d
ictio
n
a
r
y
D
th
at
g
iv
e
s
r
is
e
to
s
p
ar
s
e
s
o
l
u
ti
o
n
f
o
r
ea
ch
.
I
n
o
th
er
w
o
r
d
s
,
th
er
e
ex
i
s
ts
D,
s
u
c
h
t
h
at
s
o
l
v
i
n
g
E
q
u
a
tio
n
(
2
)
f
o
r
ea
ch
g
i
v
es
a
s
p
ar
s
e
r
ep
r
esen
tat
io
n
.
T
h
e
m
i
n
i
m
izat
io
n
p
r
o
b
lem
to
f
i
n
d
th
e
b
es
t
d
ictio
n
ar
y
f
o
r
s
p
ar
s
e
r
ep
r
esen
tatio
n
o
f
Y
i
n
t
h
e
g
i
v
e
n
s
p
ar
s
it
y
co
n
s
tr
ain
t
T
0
ca
n
b
e
r
ep
r
esen
ted
b
y
:
(
4
)
T
h
e
d
ictio
n
ar
y
i
s
tr
ain
ed
to
p
r
o
v
id
e
a
b
etter
r
ep
r
esen
tatio
n
o
f
th
e
ac
t
u
al
s
i
g
n
al
w
h
en
t
h
e
n
u
m
b
er
o
f
d
ictio
n
ar
y
ele
m
en
ts
u
s
ed
to
r
ep
r
esen
t
it
i
s
les
s
th
a
n
o
r
eq
u
al
to
T
0
.
Var
io
u
s
al
g
o
r
ith
m
s
h
a
v
e
b
ee
n
d
ev
elo
p
ed
to
tr
ain
o
v
er
co
m
p
lete
d
ictio
n
ar
i
es
f
o
r
s
p
ar
s
e
s
ig
n
al
r
ep
r
ese
n
ta
tio
n
.
T
h
e
K
-
SVD
al
g
o
r
ith
m
[
1
4
]
is
v
er
y
e
f
f
icien
t
an
d
it
w
o
r
k
s
w
ell
w
i
th
d
i
f
f
er
en
t
s
p
ar
s
e
ap
p
r
o
x
i
m
atio
n
alg
o
r
ith
m
.
K
-
SV
D
alg
o
r
it
h
m
i
te
r
ativ
el
y
u
p
d
ates
t
h
e
d
ictio
n
ar
y
ato
m
s
to
b
etter
f
it
t
h
e
d
ata.
I
n
t
h
i
s
p
ap
er
,
w
e
f
o
c
u
s
o
n
K
-
S
VD
al
g
o
r
ith
m
t
o
tr
ai
n
th
e
d
ictio
n
ar
y
a
n
d
OM
P
alg
o
r
ith
m
f
o
r
s
p
ar
s
e
r
ep
r
esen
tatio
n
.
(
a)
E
n
co
d
er
(
b
)
Dec
o
d
e
r
Fig
u
r
e
1
.
Deta
iled
b
lo
ck
d
iag
r
a
m
o
f
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Dictio
n
a
r
y
b
a
s
ed
I
ma
g
e
C
o
mp
r
ess
io
n
via
S
p
a
r
s
e
R
ep
r
esen
ta
tio
n
(
A
r
a
b
in
d
a
S
a
h
o
o
)
1967
3.
P
RO
P
O
SE
D
M
E
T
H
O
D
T
h
e
p
r
o
p
o
s
ed
im
ag
e
co
m
p
r
ess
io
n
m
et
h
o
d
co
n
s
is
ts
o
f
f
o
u
r
m
ai
n
p
r
o
ce
s
s
es:
i
n
tr
a
p
r
ed
ictio
n
,
d
ictio
n
ar
y
tr
ain
in
g
,
s
p
ar
s
e
r
ep
r
esen
tatio
n
,
a
n
d
co
d
in
g
.
T
h
e
b
lo
ck
d
iag
r
a
m
o
f
th
e
p
r
o
p
o
s
ed
co
m
p
r
ess
io
n
m
et
h
o
d
is
s
h
o
w
n
i
n
Fi
g
u
r
e
1.
3
.
1
.
I
ntr
a
P
re
dict
io
n
I
n
tr
a
p
r
ed
ictio
n
[
8
]
ex
p
lo
its
s
p
atial
co
r
r
elatio
n
w
it
h
in
o
n
e
i
m
ag
e.
I
n
t
h
is
p
r
o
ce
s
s
,
t
h
e
c
u
r
r
en
t
b
lo
ck
is
p
r
ed
icted
b
y
u
s
in
g
t
h
e
b
o
u
n
d
ar
y
p
i
x
el
v
alu
e
s
o
f
p
r
ev
io
u
s
l
y
r
ec
o
n
s
tr
u
cted
n
e
ig
h
b
o
r
in
g
b
lo
ck
s
.
A
s
s
h
o
w
n
i
n
th
e
Fi
g
u
r
e
2
,
th
e
b
o
u
n
d
ar
y
p
ix
el
v
al
u
es
o
f
n
eig
h
b
o
r
in
g
b
lo
ck
s
as
s
h
o
w
n
i
n
s
h
ad
ed
b
o
x
es
ar
e
co
p
ied
in
to
th
e
cu
r
r
en
t
b
lo
ck
p
i
x
el
s
a
lo
n
g
a
s
p
ec
if
ied
d
ir
ec
tio
n
i
n
d
icate
d
b
y
th
e
m
o
d
e.
E
ig
h
t
d
ir
ec
tio
n
al
m
o
d
es
a
n
d
a
DC
m
o
d
e,
w
h
ich
i
s
al
m
o
s
t sa
m
e
a
s
th
e
n
in
e
i
n
tr
a
-
p
r
ed
ictio
n
m
o
d
es e
m
p
lo
y
ed
in
H.
2
6
4
s
tan
d
a
r
d
[
9
]
.
(
a)
(
b
)
Fig
u
r
e
2
.
(
a)
I
n
tr
a
Pre
d
ictio
n
,
(
b
)
8
-
Dir
ec
tio
n
al
m
o
d
e
+
DC
Mo
d
e
I
n
t
h
is
p
r
o
p
o
s
ed
m
eth
o
d
,
t
h
e
i
m
a
g
e
i
s
d
i
v
id
ed
in
to
b
lo
ck
s
o
f
s
ize
8
x
8
.
T
h
e
n
i
n
e
i
n
tr
a
-
p
r
ed
ictio
n
m
o
d
e
s
(
0
-
8
)
ar
e
ap
p
lied
o
v
er
ea
ch
8
x
8
b
lo
ck
a
n
d
r
esid
u
al
er
r
o
r
is
ca
lcu
lated
f
o
r
ea
ch
p
r
ed
ictio
n
m
o
d
e.
T
h
e
r
esid
u
al
er
r
o
r
is
t
h
e
d
i
f
f
er
en
c
e
b
et
w
ee
n
t
h
e
p
i
x
el
v
al
u
e
o
f
t
h
e
c
u
r
r
en
t
b
lo
ck
an
d
p
i
x
el
v
al
u
e
o
f
t
h
e
p
r
ed
icted
b
lo
ck
.
T
h
e
b
est
p
r
ed
ictio
n
m
o
d
e
f
o
r
a
cu
r
r
en
t
b
lo
ck
i
s
s
el
ec
ted
b
ased
o
n
t
h
e
m
in
i
m
u
m
r
esid
u
al
er
r
o
r
.
T
h
e
m
o
d
e
n
u
m
b
er
M
an
d
th
e
r
esid
u
al
er
r
o
r
b
lo
ck
ar
e
tr
an
s
m
itted
f
o
r
en
co
d
in
g
.
3
.
2
.
Dict
io
na
ry
T
ra
ini
ng
T
h
e
d
ictio
n
ar
y
i
s
tr
ain
ed
o
f
f
-
l
in
e
u
s
i
n
g
p
r
ed
ictio
n
r
esid
u
al
s
a
m
p
les
r
es
u
lt
in
g
f
r
o
m
a
w
id
e
v
ar
iet
y
o
f
i
m
a
g
es.
W
e
f
ir
s
t
d
i
v
id
e
d
if
f
er
en
t
i
m
a
g
es
i
n
to
b
lo
ck
s
o
f
s
ize
8
x
8
an
d
t
h
en
9
in
tr
a
-
p
r
ed
icti
o
n
m
o
d
es
(
0
-
8
)
ar
e
ap
p
lied
o
v
er
ea
ch
8
x
8
b
lo
ck
.
T
h
e
p
r
e
d
icted
b
lo
ck
is
s
u
b
tr
ac
ted
f
r
o
m
th
e
c
u
r
r
en
t
b
lo
ck
t
o
g
en
er
ate
r
esid
u
a
l
b
lo
ck
s
.
A
s
et
o
f
8
×
8
p
r
ed
ict
io
n
r
esid
u
al
b
lo
ck
s
f
o
r
d
if
f
er
en
t
m
o
d
es
ar
e
s
elec
ted
to
tr
ain
th
e
d
ictio
n
ar
y
.
T
o
tr
ain
t
h
e
d
ictio
n
ar
y
w
e
e
m
p
l
o
y
ed
K
-
SV
D
alg
o
r
it
h
m
[
1
4
]
.
Du
r
in
g
d
ictio
n
ar
y
tr
ai
n
in
g
,
i
n
ea
ch
i
ter
atio
n
,
K
-
SVD
alg
o
r
it
h
m
u
p
d
ates
th
e
d
ictio
n
ar
y
ele
m
e
n
ts
b
y
o
p
ti
m
iz
in
g
m
i
n
i
m
izatio
n
p
r
o
b
lem
g
iv
en
in
E
q
u
at
io
n
(
4
)
.
K
-
SV
D
alg
o
r
it
h
m
iter
ativ
e
l
y
u
p
d
ates
th
e
d
ictio
n
ar
y
ele
m
en
ts
to
b
etter
f
it
th
e
d
ata
an
d
a
f
t
er
ce
r
tain
iter
a
tio
n
,
an
u
p
d
ated
d
ictio
n
ar
y
is
r
e
s
u
l
ted
.
T
h
is
u
p
d
ated
d
ictio
n
ar
y
i
s
u
s
ed
f
o
r
s
p
ar
s
e
r
ep
r
esen
tati
o
n
o
f
th
e
r
esid
u
a
l
b
lo
ck
d
u
r
in
g
i
m
a
g
e
co
d
in
g
.
3
.
3
.
Sp
a
rse
Repre
s
ent
a
t
io
n
OM
P
alg
o
r
ith
m
i
s
e
m
p
lo
y
ed
f
o
r
th
e
s
p
ar
s
e
r
ep
r
esen
tatio
n
o
f
r
esid
u
al
b
lo
ck
.
OM
P
alg
o
r
ith
m
s
e
lect
s
th
e
ap
p
r
o
p
r
iate
d
ictio
n
ar
y
ele
m
en
t
to
r
ep
r
esen
t
ea
ch
r
esid
u
al
i
m
a
g
e
b
lo
ck
.
I
n
ea
ch
iter
ati
o
n
,
OM
P
s
elec
ts
th
e
b
est
li
n
ea
r
co
m
b
i
n
atio
n
o
f
d
i
ctio
n
ar
y
e
le
m
e
n
t
s
b
y
m
in
i
m
iz
in
g
E
q
u
atio
n
(
3
)
.
T
h
e
s
a
m
e
p
r
o
ce
s
s
is
co
n
ti
n
u
e
d
an
d
t
h
e
al
g
o
r
ith
m
ter
m
i
n
ate
s
w
h
e
n
t
h
e
r
es
id
u
al
er
r
o
r
o
f
th
e
r
ec
o
n
s
tr
u
c
ted
s
i
g
n
al
i
s
eq
u
al
to
o
r
les
s
t
h
a
n
a
s
p
ec
if
ied
v
al
u
e.
Ho
w
e
v
er
,
th
e
n
u
m
b
er
o
f
OM
P
iter
atio
n
s
m
a
y
n
o
t
ex
ce
ed
T
0
.
On
ce
th
e
al
g
o
r
ith
m
ter
m
i
n
ate
s
th
e
co
ef
f
icie
n
ts
C
a
n
d
in
d
ices
P
o
f
ap
p
r
o
p
r
iate
d
ictio
n
ar
y
ele
m
en
t is tr
an
s
m
itted
f
o
r
en
co
d
i
n
g
.
3
.
4
.
Co
din
g
Af
ter
s
p
ar
s
e
r
ep
r
ese
n
tatio
n
,
th
e
co
ef
f
icie
n
t
s
C
a
n
d
in
d
i
ce
s
P
o
f
d
ictio
n
ar
y
ele
m
e
n
t
,
an
d
th
e
p
r
ed
ictio
n
m
o
d
e
M
ar
e
en
co
d
ed
[
2
2
]
.
T
h
e
co
ef
f
icie
n
ts
ar
e
u
n
i
f
o
r
m
l
y
q
u
an
tized
f
o
llo
w
ed
b
y
e
n
tr
o
p
y
co
d
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
4
,
A
u
g
u
s
t
2017
:
1
9
6
4
–
1
9
7
2
1968
T
h
e
in
d
ices
ar
e
en
co
d
ed
w
it
h
f
i
x
ed
len
g
th
co
d
es
w
h
o
s
e
s
ize
s
ar
e
lo
g
2
k
,
w
h
er
e
k
is
t
h
e
n
u
m
b
er
o
f
d
ictio
n
ar
y
ele
m
e
n
ts
.
T
h
e
p
r
ed
ictio
n
m
o
d
e
n
u
m
b
er
s
ar
e
en
co
d
ed
w
it
h
f
i
x
ed
le
n
g
th
co
d
es
w
h
o
s
e
s
i
ze
s
ar
e
4
b
it
s
.
T
h
e
en
co
d
er
o
f
th
e
p
r
o
p
o
s
ed
c
o
m
p
r
ess
io
n
m
et
h
o
d
is
s
h
o
w
n
in
Fi
g
u
r
e
1
(
a)
.
Sin
ce
t
h
e
d
ictio
n
ar
y
is
s
h
ar
e
d
at
b
o
th
en
co
d
er
an
d
d
ec
o
d
er
,
o
n
l
y
in
tr
a
-
p
r
ed
ictio
n
m
o
d
e
n
u
m
b
er
,
in
d
ices
,
an
d
co
ef
f
icien
ts
o
f
th
e
d
ictio
n
ar
y
ele
m
e
n
ts
ar
e
tr
an
s
m
i
tted
to
th
e
d
ec
o
d
e
r
.
T
h
e
d
e
co
d
er
g
en
er
ates
th
e
r
esid
u
al
b
lo
ck
f
r
o
m
t
h
e
k
n
o
w
led
g
e
o
f
d
ictio
n
ar
y
,
co
ef
f
ici
en
ts
,
a
n
d
i
n
d
ices.
T
h
e
d
ec
o
d
er
th
e
n
p
r
ed
icts
t
h
e
b
lo
ck
b
ased
o
n
m
o
d
e
n
u
m
b
er
an
d
co
m
b
in
es
w
i
th
t
h
e
r
esid
u
al
b
lo
ck
to
r
ec
o
n
s
tr
u
ct
th
e
b
lo
ck
.
T
h
e
d
ec
o
d
er
o
f
th
e
p
r
o
p
o
s
ed
co
m
p
r
ess
io
n
m
et
h
o
d
is
s
h
o
w
n
i
n
Fi
g
u
r
e
1
(
b
)
.
4.
E
XP
E
R
I
M
E
NT
A
L
RE
SUL
T
S
I
n
th
i
s
s
ec
tio
n
,
w
e
co
n
d
u
cte
d
s
ev
er
al
ex
p
er
i
m
e
n
ts
i
n
o
r
d
er
to
ev
alu
ate
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
p
r
o
p
o
s
ed
co
m
p
r
es
s
io
n
m
e
th
o
d
.
T
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
i
s
ap
p
lied
o
v
er
s
e
v
er
al
i
m
a
g
es
.
T
h
e
co
m
p
r
ess
io
n
ef
f
icien
c
y
a
n
d
q
u
alit
y
o
f
t
h
e
r
ec
o
n
s
tr
u
cted
i
m
ag
e
ar
e
c
o
m
p
ar
ed
w
it
h
s
e
v
er
al
o
th
er
co
m
p
eti
tiv
e
i
m
a
g
e
co
m
p
r
es
s
io
n
tec
h
n
iq
u
es.
4
.
1
.
I
ntr
a
P
re
dict
io
n
I
n
t
h
e
e
x
p
er
i
m
e
n
t,
w
e
u
s
ed
1
0
0
i
m
a
g
es
f
r
o
m
t
h
e
B
er
k
ele
y
s
e
g
m
e
n
tat
io
n
d
atab
ase
a
s
o
u
r
tr
a
in
i
n
g
s
et.
Nin
e
m
o
d
es
o
f
8
x
8
in
tr
a
p
r
ed
ictio
n
ar
e
ap
p
lied
an
d
in
tr
a
-
p
r
ed
ictio
n
r
esid
u
al
f
o
r
ea
ch
i
m
ag
e
is
g
e
n
er
ated
.
A
s
et
o
f
4
5
0
0
0
b
lo
ck
s
o
f
s
ize
8
x
8
ar
e
r
an
d
o
m
l
y
s
elec
ted
f
r
o
m
r
e
s
id
u
al
i
m
ag
e
s
to
tr
ai
n
a
d
ictio
n
ar
y
.
W
e
s
elec
ted
5
0
0
0
r
esid
u
al
b
lo
ck
s
f
r
o
m
ea
c
h
n
in
e
m
o
d
e
s
to
f
o
r
m
o
u
r
lear
n
in
g
s
et.
E
x
a
m
p
le
s
o
f
i
n
tr
a
-
p
r
ed
icted
i
m
a
g
e
a
n
d
r
esid
u
al
i
m
a
g
e
ar
e
d
e
m
o
n
s
tr
ated
i
n
Fig
u
r
e
3
.
A
r
an
d
o
m
co
llectio
n
o
f
s
u
ch
tr
ain
i
n
g
b
lo
ck
s
f
o
r
d
if
f
er
e
n
t
m
o
d
e
is
s
h
o
w
n
in
Fi
g
u
r
e
4
.
(
a)
Or
ig
in
al
I
m
a
g
e
(
b
)
P
r
ed
icted
I
m
a
g
e
(
c)
R
esid
u
al
I
m
a
g
e
Fig
u
r
e
3
.
I
n
tr
a
p
r
ed
ictio
n
o
f
s
ec
o
n
d
i
m
ag
e
(
a)
Mo
d
e
3
(
b
)
Mo
d
e
6
(
c)
Mo
d
e
8
Fig
u
r
e
4
.
T
r
ain
in
g
b
lo
ck
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Dictio
n
a
r
y
b
a
s
ed
I
ma
g
e
C
o
mp
r
ess
io
n
via
S
p
a
r
s
e
R
ep
r
esen
ta
tio
n
(
A
r
a
b
in
d
a
S
a
h
o
o
)
1969
4
.
2
.
K
-
SVD
Dict
io
na
r
y
W
e
u
s
ed
4
5
0
0
0
r
esid
u
al
b
lo
ck
s
o
b
tai
n
ed
f
r
o
m
i
n
tr
a
-
p
r
ed
ictio
n
a
s
o
u
r
lear
n
i
n
g
s
et.
T
h
e
K
-
SV
D
alg
o
r
ith
m
i
s
ap
p
lied
to
tr
ain
an
o
v
er
-
co
m
p
lete
d
ictio
n
ar
y
o
f
s
ize
6
4
x
5
1
2
,
w
h
er
e
t
h
e
n
u
m
b
er
o
f
r
o
w
s
64
r
ep
r
esen
ts
th
e
n
u
m
b
er
o
f
p
ix
els
i
n
a
b
lo
ck
,
an
d
th
e
n
u
m
b
er
o
f
co
lu
m
n
s
5
1
2
r
ep
r
es
en
ts
t
h
e
n
u
m
b
er
o
f
d
ictio
n
ar
y
ele
m
en
t
s
.
I
n
th
e
tr
a
in
i
n
g
p
r
o
ce
s
s
,
1
0
0
n
u
m
b
er
o
f
K
-
SV
D
iter
atio
n
s
w
a
s
s
et
as
t
h
e
p
r
i
m
ar
y
s
to
p
p
in
g
cr
iter
io
n
.
Sp
ar
s
it
y
co
n
s
t
r
ain
t
(
T
0
=4
)
w
as
s
et
as
an
o
t
h
er
s
to
p
p
in
g
cr
iter
io
n
.
E
x
a
m
p
le
o
f
a
t
r
ain
ed
d
ic
tio
n
ar
y
is
d
em
o
n
s
tr
ated
i
n
Fi
g
u
r
e
5
.
Fig
u
r
e
5
.
E
x
a
m
p
le
o
f
d
ictio
n
a
r
y
tr
ai
n
ed
o
n
8
x
8
r
esid
u
al
b
lo
c
k
s
4
.
3
.
I
m
a
g
e
Co
m
press
io
n
T
h
e
K
-
SV
D
Dict
io
n
ar
y
r
es
u
lt
in
g
f
r
o
m
p
r
ed
ictio
n
r
e
s
id
u
al
s
is
u
s
ed
f
o
r
s
p
ar
s
e
r
ep
r
esen
tati
o
n
.
I
n
t
h
e
ex
p
er
i
m
e
n
t,
th
e
r
esid
u
al
er
r
o
r
(
δ=0
.
2
)
an
d
th
e
n
u
m
b
er
o
f
OM
P
iter
atio
n
s
eq
u
al
to
4
w
a
s
s
et
as
s
to
p
p
in
g
cr
iter
io
n
o
f
OM
P
al
g
o
r
ith
m
.
T
h
e
OM
P
alg
o
r
ith
m
s
to
p
s
w
h
en
t
h
e
r
esid
u
al
er
r
o
r
is
les
s
t
h
an
o
r
eq
u
al
to
0
.
2
,
o
th
er
w
is
e
it
s
to
p
s
a
f
ter
4
ite
r
atio
n
s
.
I
n
m
ax
i
m
u
m
,
f
o
u
r
d
ictio
n
ar
y
ele
m
en
t
s
ar
e
r
eq
u
ir
ed
to
en
co
d
e
ea
c
h
r
esid
u
al
b
lo
ck
.
(
a)
(
b
)
(
c)
(
d
)
Fig
u
r
e
6
.
R
ate
-
d
is
to
r
tio
n
cu
r
v
es o
f
d
if
f
er
en
t
m
et
h
o
d
s
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
0
.
9
1
1
.
1
1
.
2
22
24
26
28
30
32
34
36
38
40
B
a
r
b
a
r
a
B
i
t
-
r
a
t
e
(
b
p
p
)
P
S
N
R
(
d
B
)
J
P
E
G
J
P
E
G
2
0
0
0
P
r
o
p
o
s
e
d
M
e
t
h
o
d
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
0
.
9
1
1
.
1
1
.
2
28
30
32
34
36
38
40
42
L
e
n
a
B
i
t
-
r
a
t
e
(
b
p
p
)
P
S
N
R
(
d
B
)
J
P
E
G
J
P
E
G
2
0
0
0
P
r
o
p
o
s
e
d
M
e
t
h
o
d
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
0
.
9
1
1
.
1
1
.
2
22
23
24
25
26
27
28
29
30
31
32
33
B
a
b
o
o
n
B
i
t
-
r
a
t
e
(
b
p
p
)
P
S
N
R
(
d
B
)
J
P
E
G
J
P
E
G
2
0
0
0
P
r
o
p
o
s
e
d
M
e
t
h
o
d
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
0
.
9
1
1
.
1
1
.
2
28
30
32
34
36
38
40
P
e
p
p
e
r
s
B
i
t
-
r
a
t
e
(
b
p
p
)
P
S
N
R
(
d
B
)
J
P
E
G
J
P
E
G
2
0
0
0
P
r
o
p
o
s
e
d
M
e
t
h
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
4
,
A
u
g
u
s
t
2017
:
1
9
6
4
–
1
9
7
2
1970
I
n
o
r
d
er
to
ev
alu
ate
o
u
r
p
r
o
p
o
s
ed
co
m
p
r
es
s
io
n
m
et
h
o
d
,
w
e
co
m
p
ar
e
o
u
r
m
et
h
o
d
w
it
h
J
P
E
G
an
d
J
P
E
G2
0
0
0
.
A
ll
ex
p
er
i
m
en
t
s
ar
e
p
er
f
o
r
m
ed
u
s
i
n
g
M
A
T
L
A
B
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
p
r
o
p
o
s
ed
m
et
h
o
d
is
ev
alu
a
ted
b
y
tak
in
g
d
if
f
er
e
n
t
s
tan
d
ar
d
test
i
m
a
g
es.
T
h
e
q
u
an
t
itati
v
e
ev
a
lu
atio
n
o
f
o
u
r
p
r
o
p
o
s
ed
m
et
h
o
d
is
ac
co
m
p
li
s
h
ed
u
s
in
g
t
w
o
i
m
ag
e
q
u
alit
y
m
etr
ics
:
P
SNR
an
d
SS
I
M
(
S
tr
u
ctu
r
al
Si
m
ilar
it
y
Me
tr
ic)
[
2
3
]
.
Fig
u
r
e
6
s
h
o
w
s
co
m
p
ar
is
o
n
o
f
r
ate
-
d
i
s
t
o
r
tio
n
cu
r
v
e
f
o
r
4
s
ta
n
d
ar
d
test
i
m
a
g
es.
T
ab
le
1
s
h
o
w
s
P
S
NR
co
m
p
ar
i
s
o
n
f
o
r
d
if
f
er
e
n
t
i
m
a
g
es
at
t
w
o
d
i
f
f
er
en
t
b
it
-
r
ates.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
y
ield
s
ar
o
u
n
d
P
SNR
g
ai
n
o
f
3
d
B
co
m
p
ar
ed
to
J
P
E
G,
an
d
P
SNR
g
ain
o
f
0
.
3
d
B
c
o
m
p
ar
ed
to
J
P
E
G
2
0
0
0
.
T
h
e
p
er
f
o
r
m
an
ce
i
n
ter
m
s
o
f
av
er
ag
e
P
S
NR
a
n
d
SS
I
M
ar
e
s
h
o
w
n
i
n
T
ab
le
2
.
T
h
e
r
esu
lts
ar
e
av
er
a
g
ed
o
v
er
9
s
tan
d
ar
d
test
i
m
a
g
es
a
n
d
b
est
r
esu
lt
s
ar
e
b
o
ld
ed
.
Su
b
j
ec
tiv
e
a
s
s
e
s
s
m
en
t
o
f
o
n
e
i
m
a
g
e
i
s
s
h
o
w
n
i
n
Fi
g
u
r
e
7
at
b
it
-
r
ate
0
.
2
.
T
h
e
r
esu
lt
s
s
h
o
w
th
at
o
u
r
p
r
o
p
o
s
ed
m
et
h
o
d
o
u
tp
er
f
o
r
m
s
J
P
E
G
an
d
J
P
E
G2
0
0
0
in
ter
m
s
o
f
all
q
u
alit
y
m
etr
ics,
i
n
cl
u
d
in
g
P
SNR
an
d
SS
I
M.
T
ab
le
1.
P
SNR
(
d
B
)
c
o
m
p
ar
is
o
n
o
f
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
w
it
h
J
P
G
an
d
J
P
E
G2
0
0
0
f
o
r
a
s
et
o
f
9
test
i
m
a
g
es a
t
t
w
o
d
if
f
er
e
n
t b
it
-
r
ate
s
.
B
est r
esu
lt
s
ar
e
b
o
ld
ed
P
S
N
R
(
d
B
)
a
t
b
i
t
-
r
a
t
e
0
.
2
b
p
p
P
S
N
R
(
d
B
)
a
t
b
i
t
-
r
a
t
e
1
b
p
p
I
mag
e
s
JP
EG
JP
EG
2
0
0
0
P
r
o
p
o
se
d
me
t
h
o
d
JP
EG
JP
EG
2
0
0
0
P
r
o
p
o
se
d
me
t
h
o
d
B
a
r
b
a
r
a
2
3
.
8
1
2
7
.
2
9
2
7
.
6
0
3
4
.
4
6
3
7
.
1
5
3
7
.
5
2
S
a
i
l
b
o
a
t
2
6
.
5
6
2
9
.
1
2
2
9
.
5
3
3
4
.
4
4
3
6
.
8
1
3
7
.
0
1
B
a
b
o
o
n
2
2
.
4
8
2
5
.
2
4
2
5
.
4
5
2
7
.
9
8
3
0
.
6
5
3
0
.
8
1
c
o
u
p
l
e
2
5
.
7
2
2
8
.
4
8
2
8
.
8
0
3
4
.
2
5
3
6
.
7
9
3
7
.
1
0
H
i
l
l
2
6
.
7
8
2
9
.
8
8
3
0
.
2
1
3
3
.
6
2
3
6
.
4
2
3
6
.
8
4
Je
t
p
l
a
n
e
2
9
.
5
6
3
1
.
9
2
3
2
.
2
8
3
9
.
6
5
4
1
.
8
8
4
2
.
2
0
L
e
n
a
2
9
.
4
3
3
3
.
0
2
3
3
.
4
6
3
7
.
8
8
4
0
.
4
0
4
0
.
8
2
L
i
g
h
t
h
o
u
se
2
5
.
8
2
2
8
.
3
9
2
8
.
8
2
3
5
.
5
0
3
8
.
4
2
3
8
.
8
3
P
e
p
p
e
r
s
2
9
.
9
8
3
2
.
4
9
3
2
.
8
8
3
7
.
5
5
3
8
.
3
8
3
8
.
7
2
A
v
e
r
a
g
e
2
6
.
6
8
2
9
.
5
4
2
9
.
8
9
3
5
.
0
3
3
7
.
4
3
3
7
.
7
6
(
a)
Or
ig
in
al
i
m
a
g
e
(
b
)
J
P
E
G
(
c)
J
P
E
G2
0
0
0
(
d
)
P
r
o
p
o
s
ed
m
et
h
o
d
Fig
u
r
e
7
.
Su
b
j
ec
tiv
e
ass
e
s
s
m
e
n
t
T
ab
le
2
.
P
er
f
o
r
m
a
n
ce
co
m
p
ar
i
s
o
n
in
ter
m
s
o
f
t
w
o
i
m
a
g
e
q
u
a
lit
y
m
etr
ics,
P
SNR
(
d
B
)
an
d
SS
I
M
at
s
ix
d
if
f
er
e
n
t b
it
-
r
ate
s
.
T
h
e
r
esu
lts
ar
e
av
er
ag
ed
o
v
er
9
test
i
m
a
g
e
s
.
B
est r
esu
lt
s
ar
e
b
o
ld
ed
Q
u
a
l
i
t
y
M
e
t
r
i
c
JP
EG
JP
EG
2
0
0
0
P
r
o
p
o
se
d
me
t
h
o
d
JP
EG
JP
EG
2
0
0
0
P
r
o
p
o
se
d
me
t
h
o
d
B
i
t
-
r
a
t
e
0
.
2
b
p
p
Bit
-
r
a
t
e
0
.
4
b
p
p
P
S
N
R
2
6
.
6
8
2
9
.
5
4
2
9
.
8
9
2
8
.
7
8
3
1
.
8
2
3
2
.
1
4
S
S
I
M
0
.
7
4
8
0
.
7
7
1
0
.
7
8
2
0
.
7
6
2
0
.
8
4
4
0
.
8
6
1
Bit
-
r
a
t
e
0
.
6
b
p
p
Bit
-
r
a
t
e
0
.
8
b
p
p
P
S
N
R
3
1
.
1
2
3
3
.
8
2
3
4
.
2
2
3
3
.
4
0
3
6
.
4
2
3
6
.
7
4
S
S
I
M
0
.
8
2
4
0
.
9
0
6
0
.
9
1
5
0
.
8
7
4
0
.
9
2
1
0
.
9
3
1
Bit
-
r
a
t
e
1
b
p
p
Bit
-
r
a
t
e
1
.
2
b
p
p
P
S
N
R
3
5
.
0
3
3
7
.
4
3
3
7
.
7
6
3
6
.
4
3
3
8
.
5
5
3
8
.
8
7
S
S
I
M
0
.
9
0
9
0
.
9
4
1
0
.
9
4
9
0
.
9
1
8
0
.
9
5
1
0
.
9
5
4
50
100
150
200
250
300
350
400
450
500
50
100
150
200
250
300
350
400
450
500
50
100
150
200
250
300
350
400
450
500
50
100
150
200
250
300
350
400
450
500
50
100
150
200
250
300
350
400
450
500
50
100
150
200
250
300
350
400
450
500
50
100
150
200
250
300
350
400
450
500
50
100
150
200
250
300
350
400
450
500
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Dictio
n
a
r
y
b
a
s
ed
I
ma
g
e
C
o
mp
r
ess
io
n
via
S
p
a
r
s
e
R
ep
r
esen
ta
tio
n
(
A
r
a
b
in
d
a
S
a
h
o
o
)
1971
5.
CO
NCLU
SI
O
N
I
n
th
i
s
p
ap
er
,
w
e
p
r
ese
n
ted
a
d
ictio
n
ar
y
b
ased
in
tr
a
-
p
r
ed
ictio
n
f
r
a
m
e
w
o
r
k
f
o
r
i
m
a
g
e
co
m
p
r
e
s
s
io
n
.
K
-
SV
D
alg
o
r
it
h
m
is
u
s
ed
i
n
o
r
d
er
to
tr
ain
a
d
ictio
n
ar
y
.
W
e
tr
ain
ed
th
e
d
ictio
n
ar
y
w
it
h
a
v
ar
iet
y
o
f
r
esid
u
a
l
b
lo
ck
s
o
b
tain
ed
f
r
o
m
in
tr
a
-
p
r
ed
ictio
n
an
d
th
e
n
u
s
e
d
t
h
i
s
d
ictio
n
ar
y
f
o
r
s
p
ar
s
e
r
ep
r
esen
t
atio
n
o
f
an
i
m
ag
e.
OM
P
alg
o
r
ith
m
,
f
i
x
ed
len
g
t
h
co
d
in
g
,
an
d
en
tr
o
p
y
co
d
in
g
h
av
e
e
m
p
lo
y
ed
f
o
r
en
co
d
in
g
.
Dif
f
er
e
n
t
co
d
in
g
r
esu
lt
s
b
ased
o
n
a
s
et
o
f
test
i
m
ag
e
s
ar
e
p
r
esen
ted
to
co
m
p
a
r
e
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
w
it
h
th
e
ex
i
s
ti
n
g
m
et
h
o
d
s
.
E
x
p
er
i
m
en
tal
r
es
u
lt s
h
o
w
s
t
h
at
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
o
u
tp
er
f
o
r
m
s
J
P
E
G
an
d
J
P
E
G2
0
0
0
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
e
au
th
o
r
s
ar
e
g
r
atef
u
l to
Dr
.
Gag
an
R
at
h
f
o
r
h
i
s
s
u
p
p
o
r
t a
n
d
g
u
id
an
ce
t
h
r
o
u
g
h
o
u
t t
h
is
w
o
r
k
.
RE
F
E
R
E
NC
E
S
[1
]
S
a
lo
m
o
n
,
“
Da
ta
Co
m
p
re
ss
io
n
:
T
h
e
Co
mp
lete
Refe
re
n
c
e
”
,
t
h
ir
d
e
d
.
,
S
p
rin
g
e
r,
Ne
w
Yo
rk
,
2
0
0
4
.
[2
]
W
.
P
e
n
n
e
b
a
k
e
r,
J.
M
it
c
h
e
l
l,
“
J
PE
G S
ti
ll
Im
a
g
e
D
a
ta
Co
mp
re
ss
io
n
S
ta
n
d
a
rd
”
,
Klu
w
e
r
A
c
a
d
e
m
ic P
u
b
li
sh
e
rs
,
1
9
9
3
.
[3
]
C.
Ch
risto
p
o
u
l
o
s,
A
.
S
k
o
d
ra
s,
T
.
Eb
ra
h
im
i,
“
T
h
e
JP
EG
2
0
0
0
S
t
il
l
Im
a
g
e
Co
d
i
n
g
S
y
ste
m
:
A
n
O
v
e
rv
ie
w
”
,
IEE
E
T
ra
n
s.
C
o
n
s
u
m.
El
e
c
tro
n
,
4
6
(
2
0
0
0
)
1
1
0
3
-
1
1
2
7
.
[4
]
A
.
En
n
a
c
iri
,
M
.
Err
it
a
li
,
M
.
M
a
b
ro
u
k
i,
J.
Be
n
g
o
u
rra
m
,
“
Co
m
p
a
ra
ti
v
e
S
tu
d
y
o
f
Wav
e
let
I
m
a
g
e
Co
m
p
re
ss
io
n
:
JP
EG
2
0
0
0
S
tan
d
a
rt
”
,
T
EL
KOM
N
IKA
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
E
lec
trica
l
En
g
in
e
e
rin
g
,
v
o
l.
1
6
,
n
o
.
1
,
p
p
.
8
3
-
9
0
,
Oc
t.
2
0
1
5
.
[5
]
R.
Ch
o
u
ra
siy
a
,
A
.
S
h
riv
a
sta
v
a
,
“
A
S
tu
d
y
o
f
I
m
a
g
e
Co
m
p
re
ss
io
n
Ba
se
d
T
ra
n
s
m
issio
n
A
l
g
o
rit
h
m
u
sin
g
S
P
IHT
f
o
r
L
o
w
Bit
Ra
te
A
p
p
li
c
a
ti
o
n
”
,
Bu
ll
e
ti
n
o
f
El
e
c
trica
l
E
n
g
i
n
e
e
rin
g
a
n
d
In
fo
rm
a
t
ics
,
v
o
l.
2
,
n
o
.
2
,
p
p
.
1
1
7
-
1
2
2
,
J
u
n
.
2
0
1
3
.
[6
]
D.
Na
r
m
a
d
h
a
,
K.
Ga
y
a
th
ri,
K.
T
h
il
a
g
a
v
a
th
i,
N
.
S
a
rd
a
r
Ba
sh
a
,
“
An
Op
ti
m
a
l
HSI
I
m
a
g
e
Co
m
p
re
ss
i
o
n
u
si
n
g
DWT
a
n
d
C
P
”
,
I
n
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
C
o
mp
u
ter
En
g
i
n
e
e
rin
g
,
v
o
l
.
4
,
n
o
.
3
,
p
p
.
4
1
1
-
4
2
1
,
J
u
n
.
2
0
1
4
.
[7
]
M
.
S
ti
rn
e
r,
G
.
S
e
e
lma
n
n
,
“
Imp
ro
v
e
d
R
e
d
u
n
d
a
n
c
y
Red
u
c
ti
o
n
fo
r
J
PE
G
Fi
les
”
,
P
ro
c
.
o
f
P
ictu
re
Co
d
in
g
S
y
m
p
o
siu
m
(P
C
S
2
0
0
7
)
,
L
isb
o
n
,
P
o
rt
u
g
a
l,
No
v
.
7
-
9
,
2
0
0
7
.
[8
]
J.
Ya
n
g
,
B.
Yin
,
Y.
S
u
n
,
N.
Z
h
a
n
g
,
“
A
b
lo
c
k
-
M
a
tch
in
g
b
a
se
d
In
tr
a
F
ra
m
e
P
re
d
icti
o
n
H.2
6
4
/A
V
C”
,
ICM
E
,
2
0
0
6
.
[9
]
T
.
W
i
e
g
a
n
d
,
G
.
J
.
S
u
ll
iv
a
n
,
G
.
Bjø
n
teg
a
a
rd
,
A
.
L
u
th
ra
,
“
Ov
e
r
v
i
e
w
o
f
th
e
H.2
6
4
/A
V
C
V
id
e
o
C
o
d
in
g
S
tan
d
a
rd
”
,
IEE
E
T
ra
n
s.
Circ
u
it
s S
y
st.
Vi
d
e
o
T
e
c
h
n
o
l
ogy
.
,
v
o
l.
1
3
,
n
o
.
7
,
p
p
.
5
6
0
-
5
7
6
,
Ju
l
.
2
0
0
3
.
[1
0
]
J.
W
rig
h
t,
Y.
M
a
,
J.
M
a
i
ra
l,
G
.
S
a
p
iro
,
T
.
S
.
Hu
a
n
g
,
S
.
Ya
n
,
“
S
p
a
rs
e
Rep
re
se
n
ta
ti
o
n
fo
r
C
o
mp
u
te
r
Vi
sio
n
a
n
d
Pa
tt
e
rn
Rec
o
g
n
it
io
n
”
,
P
ro
c
.
IEE
E
,
v
o
l.
9
8
,
n
o
.
6
,
p
p
.
1
0
3
1
–
1
0
4
4
,
Ju
n
.
2
0
1
0
.
[1
1
]
R.
R
u
b
i
n
ste
in
,
A
.
M
.
Bru
c
k
ste
in
,
M
.
El
a
d
,
“
Dic
ti
o
n
a
rie
s
fo
r
S
p
a
r
se
Rep
re
se
n
ta
ti
o
n
M
o
d
e
li
n
g
”
,
P
r
o
c
.
IEE
E
,
v
o
l
.
9
8
,
n
o
.
6
,
p
p
.
1
0
4
5
–
1
0
5
7
,
J
u
n
e
2
0
1
0
.
[1
2
]
J.
M
a
iral
,
F
.
Ba
c
h
,
J.
P
o
n
c
e
,
a
n
d
G
.
S
a
p
iro
,
“
On
li
n
e
d
ictio
n
a
ry
lea
rn
in
g
f
o
r
sp
a
rse
c
o
d
in
g
,
”
in
P
r
o
c
.
2
6
t
h
A
n
n
u
.
In
t.
Co
n
f.
M
a
c
h
.
L
e
a
rn
i
n
g
.
,
2
0
0
9
,
p
p
.
6
8
9
-
6
9
6
.
[1
3
]
O.
Br
y
t,
M
.
El
a
d
,
“
Co
m
p
re
ss
i
o
n
o
f
F
a
c
ial
I
m
a
g
e
s
u
sin
g
th
e
K
-
S
V
D
A
lg
o
rit
h
m
”
,
J
.
Vi
s.
C
o
mm
u
n
.
Ima
g
e
Rep
re
se
n
ta
ti
o
n
.
v
o
l.
1
9
,
n
o
.
4
,
p
p
.
2
7
0
-
2
8
2
,
2
0
0
8
.
[1
4
]
M
.
A
h
a
ro
n
,
M
.
El
a
d
,
A
.
Bru
c
k
st
e
in
,
“
K
-
S
V
D:
A
n
A
lg
o
rit
h
m
f
o
r
De
sig
n
in
g
Ov
e
rc
o
m
p
lete
D
ictio
n
a
ries
f
o
r
S
p
a
rse
R
e
p
re
se
n
tatio
n
,
”
I
EE
E
T
r
a
n
s.
S
i
g
n
a
l
Pro
c
e
ss
in
g
.
,
v
o
l.
5
4
,
n
o
.
1
1
,
p
p
.
4
3
1
1
-
4
3
2
2
,
N
o
v
.
2
0
0
6
.
[1
5
]
J.
Zep
e
d
a
,
C.
G
u
il
lem
o
t,
E.
Kija
k
,
“
Im
a
g
e
Co
m
p
re
ss
io
n
u
si
n
g
S
p
a
rse
Re
p
re
se
n
tatio
n
s
a
n
d
t
h
e
Iter
a
ti
o
n
-
t
u
n
e
d
a
n
d
A
li
g
n
e
d
Dic
ti
o
n
a
ry
”
,
IEE
E
J
.
S
e
l.
T
o
p
ics
S
i
g
n
a
l
Pr
o
c
e
ss
in
g
,
v
o
l
.
5
,
n
o
.
5
,
p
p
.
1
0
6
1
-
1
0
7
3
,
S
e
p
.
2
0
1
1
.
[1
6
]
G
.
S
h
a
o
,
Y.
W
u
,
A
.
Yo
n
g
,
X
.
L
iu
,
T
.
G
u
o
,
“
F
in
g
e
rp
r
in
t
Co
m
p
re
ss
io
n
b
a
se
d
o
n
S
p
a
rse
Re
p
re
se
n
tatio
n
”
,
IE
EE
T
ra
n
s.
Im
a
g
e
Pr
o
c
e
ss
.
,
v
o
l.
2
3
,
n
o
.
2
,
p
p
.
4
8
9
-
5
0
1
,
F
e
b
.
2
0
1
4
.
[1
7
]
Jin
g
-
Ya
Zh
u
,
Zh
o
n
g
-
Yu
a
n
W
a
n
g
,
Ru
i
Z
h
o
n
g
,
S
h
e
n
-
M
i
n
g
Qu
,
“
Dic
ti
o
n
a
ry
b
a
se
d
S
u
rv
e
il
lan
c
e
Im
a
g
e
Co
m
p
re
ss
io
n
”
,
J
.
Vi
s.
C
o
mm
u
n
.
I
ma
g
e
R
.
v
o
l.
31
,
n
o
.
7
,
p
p
.
2
2
5
-
2
3
0
.
Ju
l
.
2
0
1
5
.
[1
8
]
G
.
Ra
th
,
A
.
S
a
h
o
o
,
“
A
C
o
m
p
a
ra
ti
v
e
S
tu
d
y
o
f
so
m
e
G
re
e
d
y
P
u
rsu
it
A
lg
o
rit
h
m
s
f
o
r
S
p
a
rse
Ap
p
r
o
x
im
a
ti
o
n
”
,
EUS
IPCO2
0
0
9
,
p
p
.
3
9
8
-
4
0
2
,
A
u
g
.
2
0
1
1
.
[1
9
]
S
.
M
a
ll
a
t,
Z.
Zh
a
n
g
,
“
M
a
tch
in
g
P
u
rsu
i
ts
w
it
h
T
i
m
e
F
re
q
u
e
n
c
y
Dic
ti
o
n
a
ries
”
,
IEE
E
T
ra
n
sa
c
t
io
n
s
o
n
S
ig
n
a
l
Pro
c
e
ss
in
g
,
v
o
l
.
4
1
,
n
o
.
1
2
,
p
p
.
3
3
9
7
-
3
4
1
5
,
De
c
.
1
9
9
3
.
[2
0
]
Y.C.
P
a
ti
,
R.
Re
z
a
ii
f
a
r
M,
P
.
S
.
Kri
sh
n
a
p
ra
sa
d
,
“
Or
th
o
g
o
n
a
l
M
a
tch
i
n
g
p
u
rs
u
it
:
Rec
u
rs
ive
Fu
n
c
ti
o
n
Ap
p
ro
x
im
a
ti
o
n
wit
h
A
p
p
l
ica
ti
o
n
t
o
W
a
v
e
let
D
e
c
o
mp
o
siti
o
n
”
,
In
P
ro
c
.
2
7
th
Co
n
f
.
o
n
S
ig
.
S
y
s.
a
n
d
C
o
m
p
.
,
v
o
l.
1
,
No
v
.
1
9
9
3
.
[2
1
]
G
.
Ra
th
,
C.
G
u
il
lem
o
t,
“
On
a
S
im
p
le
De
riv
a
ti
o
n
o
f
th
e
C
o
m
p
li
m
e
n
tar
y
M
a
tch
in
g
P
u
rsu
it
”
,
S
i
g
n
a
l
p
ro
c
e
ss
in
g
,
v
o
l.
9
0
,
n
o
.
2
,
p
p
.
7
02
-
7
0
6
,
F
e
b
.
2
0
1
0
.
[2
2
]
A
.
Bo
v
ik
,
Ha
n
d
b
o
o
k
o
f
Ima
g
e
a
n
d
Vi
d
e
o
Pr
o
c
e
ss
in
g
,
2
n
d
e
d
.
S
a
n
F
ra
n
c
isc
o
,
CA
,
USA
:
A
c
a
d
e
m
i
c
,
2
0
0
5
.
[2
3
]
Z.
W
a
n
g
,
A
.
C.
Bo
v
ik
,
H.
R.
S
h
e
ik
h
,
E.
P
.
S
i
m
o
n
c
e
ll
i,
“
Im
a
g
e
Qu
a
li
ty
A
ss
e
ss
m
e
n
t:
F
ro
m
Err
o
r
V
isib
i
li
ty
to
S
tru
c
tu
ra
l
S
im
il
a
rit
y
,
”
IEE
E
T
ra
n
s.
Ima
g
e
Pr
o
c
e
ss
in
g
.
,
v
o
l.
1
3
,
n
o
.
4
,
p
p
.
6
0
0
-
6
1
2
,
A
p
r.
2
0
0
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
4
,
A
u
g
u
s
t
2017
:
1
9
6
4
–
1
9
7
2
1972
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
Ara
bin
da
Sa
ho
o
.
C
u
r
r
en
t
l
y
,
h
e
is
w
o
r
k
i
n
g
as
an
Ass
i
s
tan
t
P
r
o
f
ess
o
r
in
th
e
d
ep
ar
t
m
en
t
o
f
E
lectr
o
n
ics
&
C
o
m
m
u
n
icatio
n
E
n
g
i
n
ee
r
in
g
at
I
T
E
R
,
Sik
s
h
a
„
O
‟
An
u
s
a
n
d
h
a
n
Un
i
v
er
s
i
t
y
,
B
h
u
b
an
e
s
w
ar
,
Od
is
h
a,
I
n
d
i
a.
He
h
as
co
m
p
leted
h
is
B
.
E
f
r
o
m
Utk
al
U
n
i
v
er
s
it
y
,
B
h
u
b
an
e
s
w
ar
,
Od
is
h
a,
I
n
d
ia
an
d
M.
T
ec
h
f
r
o
m
Natio
n
al
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
,
R
o
u
r
k
el
a
,
I
n
d
ia
.
His
r
esear
ch
i
n
ter
ests
i
n
clu
d
e
s
i
g
n
al
a
n
d
i
m
a
g
e
p
r
o
ce
s
s
i
n
g
,
i
m
ag
e
co
m
p
r
ess
io
n
a
n
d
s
p
ar
s
e
r
ep
r
esen
ta
tio
n
s
.
P
ra
na
t
i
Da
s
.
C
u
r
r
en
tl
y
,
sh
e
is
w
o
r
k
in
g
a
s
a
n
A
s
s
o
ciate
P
r
o
f
ess
o
r
in
t
h
e
d
ep
ar
t
m
e
n
t
o
f
E
lectr
ical
E
n
g
i
n
ee
r
i
n
g
at
I
n
d
ir
a
Gan
d
h
i
I
n
s
tit
u
te
o
f
T
ec
h
n
o
lo
g
y
,
Sar
a
n
g
a
n
A
u
to
n
o
m
o
u
s
I
n
s
tit
u
te
o
f
Go
v
er
n
m
en
t
o
f
Od
is
h
a,
I
n
d
ia.
Sh
e
h
a
s
co
m
p
leted
h
er
M.
T
ec
h
&
P
h
D
f
r
o
m
I
n
d
ian
I
n
s
tit
u
te
o
f
T
ec
h
n
o
l
o
g
y
,
K
h
ar
ag
p
u
r
,
I
n
d
ia
a
n
d
E
n
g
i
n
ee
r
i
n
g
g
r
ad
u
atio
n
f
r
o
m
Sa
m
b
alp
u
r
Un
iv
er
s
it
y
,
Od
is
h
a
,
I
n
d
ia.
Sh
e
h
as
p
u
b
lis
h
ed
n
u
m
b
er
o
f
r
esear
ch
an
d
co
n
f
er
en
ce
p
ap
er
s
in
b
o
th
n
atio
n
al
an
d
in
ter
n
at
io
n
al
f
o
r
u
m
.
Her
ar
ea
o
f
r
esear
c
h
in
ter
est
s
ar
e
I
m
a
g
e
P
r
o
ce
s
s
in
g
a
n
d
Sig
n
al
P
r
o
ce
s
s
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.