I
nte
rna
t
io
na
l J
o
urna
l o
f
Adv
a
nces in Applie
d Science
s
(
I
J
AAS)
Vo
l.
4
,
No
.
1
,
Ma
r
ch
2
0
1
5
,
p
p
.
24
~
31
I
SS
N:
2252
-
8814
24
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
J
AAS
An Enha
nced
Lo
ss
less
Colo
r F
ilt
er
Array
I
m
a
g
e Co
m
pr
ess
io
n
Ba
sed o
n P
redi
cti
v
e Adaptiv
e A
rith
m
e
tic
Co
ding
L.
M
.
V
a
ra
la
k
s
h
m
i,
R.
So
wm
iy
a
De
p
a
rtme
n
t
o
f
El
e
c
tro
n
ics
&
Co
m
m
u
n
ica
ti
o
n
En
g
in
e
e
rin
g
,
S
ri
M
a
n
a
k
u
la V
i
n
a
y
a
g
a
r
En
g
in
e
e
rin
g
C
o
ll
e
g
e
,
In
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Dec
1
1
,
2
0
1
4
R
ev
i
s
ed
Feb
10
,
2
0
1
5
A
cc
ep
ted
Feb
19
,
2
0
1
5
M
o
st
c
o
n
s
u
m
e
r
d
ig
it
a
l
c
a
m
e
ra
s
u
se
a
sin
g
le
i
m
a
g
e
li
g
h
t
se
n
so
r
w
h
ich
p
ro
v
id
e
s
c
o
lo
r
in
f
o
rm
a
ti
o
n
u
sin
g
c
o
lo
r
f
il
ter
a
rra
y
(CF
A
).
T
h
is
p
ro
v
id
e
d
a
m
o
sa
ic
i
m
a
g
e
s,
in
w
h
ich
e
a
c
h
p
ix
e
l
p
o
siti
o
n
c
o
n
tain
s
o
n
ly
o
n
e
c
o
lo
r
c
o
m
p
o
n
e
n
t
in
c
a
se
o
f
Ba
y
e
r
C
F
A
P
a
tt
e
rn
.
T
h
is
p
a
p
e
r
p
ro
d
u
c
e
d
a
CF
A
h
iera
rc
h
ica
l
p
re
d
ictio
n
sc
h
e
m
e
b
a
se
d
o
n
c
o
n
tex
t
a
d
a
p
ti
v
e
c
o
d
in
g
.
In
CF
A
h
iera
rc
h
ica
l
sc
h
e
m
e
,
th
e
g
re
e
n
p
ix
e
ls
w
e
re
su
b
d
iv
i
d
e
d
in
t
o
tw
o
se
ts
.
w
a
s
e
n
c
o
d
e
d
b
y
a
g
ra
y
sc
a
le
c
o
n
v
e
n
ti
o
n
a
l
m
e
th
o
d
a
n
d
w
a
s
p
re
d
icte
d
b
a
se
d
o
n
.
T
h
e
re
d
p
ix
e
ls
w
e
re
p
re
d
icte
d
u
si
n
g
b
o
th
th
e
se
ts
o
f
g
re
e
n
p
ix
e
ls
a
n
d
b
lu
e
p
ix
e
ls
w
e
r
e
p
re
d
icte
d
u
sin
g
re
d
a
n
d
g
re
e
n
.
T
h
e
p
re
d
icto
rs
w
e
re
d
e
sig
n
e
d
b
a
s
e
d
o
n
d
irec
ti
o
n
o
f
th
e
e
d
g
e
s
in
th
e
n
e
ig
h
b
o
r
h
o
o
d
.
Us
in
g
t
h
e
p
re
d
ict
i
o
n
i
n
f
o
rm
a
ti
o
n
,
th
e
m
a
g
n
it
u
d
e
o
f
p
re
d
icti
o
n
e
rro
r
w
a
s
a
lso
d
e
ter
m
in
e
d
a
n
d
c
o
n
t
e
x
t
a
d
a
p
ti
v
e
a
rit
h
m
e
ti
c
c
o
d
i
n
g
w
a
s
a
p
p
li
e
d
t
o
re
d
u
c
e
b
it
s.
T
h
e
sim
u
l
a
ted
re
su
lt
s
o
n
CF
A
i
m
a
g
e
s
sh
o
w
ed
th
a
t
th
e
p
ro
p
o
se
d
m
e
th
o
d
g
iv
e
s
les
s
b
it
s
p
e
r
p
ix
e
l
th
a
n
th
e
re
c
e
n
tl
y
d
e
v
e
l
o
p
e
d
CF
A
c
o
m
p
re
ss
io
n
a
lg
o
rit
h
m
s.
K
ey
w
o
r
d
:
C
o
lo
r
f
ilter
ar
r
a
y
C
o
n
te
x
t c
o
d
in
g
Hier
ar
ch
ical
p
r
ed
ictio
n
L
o
s
s
less
co
m
p
r
e
s
s
io
n
Co
p
y
rig
h
t
©
201
5
In
s
t
it
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
:
L
.
M.
Var
alak
s
h
m
i
,
Dep
ar
t
m
en
t o
f
E
lectr
o
n
ics
&
C
o
m
m
u
n
ica
tio
n
E
n
g
i
n
ee
r
in
g
,
Sri
Ma
n
a
k
u
la
V
in
a
y
ag
ar
E
n
g
i
n
ee
r
in
g
C
o
lle
g
e
,
P
u
d
u
ch
er
r
y
-
6
0
5
1
0
7
,
I
n
d
ia
.
E
m
ail:
v
ar
alak
s
h
m
i
-
1
@
y
a
h
o
o
.
co
.
in
1.
I
NT
RO
D
UCT
I
O
N
W
ith
an
i
n
cr
ea
s
i
n
g
attr
ac
ti
v
e
n
es
s
o
f
d
ig
ital
ca
m
er
as,
t
h
e
q
u
an
t
it
y
o
f
d
ig
ital
i
m
a
g
es
i
s
co
n
s
id
er
ab
l
y
g
r
o
w
i
n
g
,
a
n
d
t
h
e
r
eso
l
u
tio
n
o
f
d
ig
i
tal
i
m
a
g
es
is
also
q
u
i
ck
l
y
i
n
cr
ea
s
i
n
g
.
D
i
g
ital
ca
m
er
as
ar
e
b
ec
o
m
i
n
g
g
r
ad
u
all
y
m
o
r
e
p
o
p
u
lar
in
t
h
e
u
s
er
elec
tr
o
n
ic
s
m
ar
k
e
t.
T
o
r
ed
u
ce
th
e
co
s
t,
m
o
s
t
v
id
eo
ca
m
er
as
r
ep
ea
ted
l
y
ca
p
tu
r
e
th
e
co
lo
r
in
f
o
r
m
ati
o
n
u
s
i
n
g
a
s
i
n
g
le
ch
ar
g
e
-
c
o
u
p
led
d
ev
ice
o
r
co
m
p
le
m
e
n
tar
y
m
etal
-
o
x
id
e
-
s
e
m
ico
n
d
u
cto
r
s
e
n
s
o
r
i
m
a
g
i
n
g
p
ip
elin
e
w
it
h
th
e
r
ed
-
g
r
ee
n
-
b
l
u
e
(
R
GB
)
co
lo
r
f
ilter
ar
r
a
y
(
C
F
A
)
s
tr
u
ct
u
r
e.
T
h
e
B
ay
er
p
a
tter
n
C
F
A
s
h
o
w
n
i
n
Fig
u
r
e
1
i
s
co
m
m
o
n
l
y
u
s
ed
,
wh
ich
ca
p
tu
r
e
s
p
i
x
els
w
h
ich
co
n
tai
n
in
g
t
w
o
g
r
ee
n
,
o
n
e
r
ed
an
d
o
n
e
b
lu
e
s
a
m
p
le
(
Fig
u
r
e
1
)
.
Fig
u
r
e
1
.
B
ay
er
C
F
A
p
atter
n
[
1
]
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
A
n
E
n
h
a
n
ce
d
Lo
s
s
less
C
o
lo
r
F
ilter
A
r
r
a
y
I
ma
g
e
C
o
mp
r
ess
i
o
n
B
a
s
ed
o
n
P
r
ed
i
ctive
…
(
L.
M.
V
a
r
a
la
ksh
mi
)
25
C
o
lo
r
d
em
o
s
aic
k
i
n
g
(
C
DM
)
d
eter
m
i
n
es
m
is
s
in
g
t
w
o
co
lo
r
c
o
m
p
o
n
en
t
s
in
ea
c
h
p
ix
el
lo
ca
tio
n
o
f
th
e
C
F
A
i
m
a
g
e
to
p
r
o
d
u
ce
th
e
f
u
ll
co
lo
r
i
m
a
g
e.
I
n
a
co
n
s
er
v
ati
v
e
d
ig
ital
ca
m
er
a
p
ip
elin
e,
C
DM
is
i
n
itial
l
y
p
er
f
o
r
m
s
o
n
t
h
e
C
F
A
i
m
a
g
e,
f
o
llo
w
ed
b
y
co
m
p
r
e
s
s
io
n
o
f
d
e
m
o
s
aic
k
ed
i
m
a
g
e.
A
ca
m
er
a
’
s
i
m
a
g
e
p
r
o
ce
s
s
o
r
p
er
f
o
r
m
s
m
o
s
t
o
f
t
h
e
p
r
ep
r
o
c
ess
i
n
g
s
tep
s
s
u
c
h
a
s
w
h
ite
b
a
lan
ci
n
g
,
d
e
n
o
is
i
n
g
,
a
n
d
d
e
m
o
s
aick
i
n
g
u
s
i
n
g
th
e
r
a
w
d
ata
ca
p
tu
r
ed
b
y
C
F
A
T
h
e
d
e
m
o
s
aic
k
ed
co
lo
r
i
m
ag
e
is
th
e
n
s
a
v
ed
as
r
a
w
R
GB
i
m
ag
e
o
r
s
en
d
to
i
m
ag
e
en
co
d
er
.
Hen
ce
,
b
y
p
r
o
ce
s
s
i
n
g
o
f
th
ese
i
m
a
g
es
i
n
a
s
y
s
te
m
is
s
o
m
eti
m
es
r
ed
u
n
d
an
t
b
ec
au
s
e
th
e
C
F
A
d
ata
h
a
s
b
ee
n
alr
ea
d
y
p
r
o
ce
s
s
ed
u
s
i
n
g
t
h
e
ca
m
er
a
’
s
p
r
o
ce
s
s
o
r
.
Data
co
m
p
r
ess
io
n
is
a
n
o
t
h
er
i
m
p
o
r
tan
t
p
r
o
ce
s
s
i
n
g
s
ta
g
e
i
n
d
ig
ital
ca
m
er
as,
to
r
ed
u
ce
t
h
e
s
to
r
ag
e
lev
el,
r
ed
u
n
d
a
n
cies
ar
e
r
e
m
o
v
ed
f
r
o
m
t
h
e
d
ata.
C
o
m
p
r
es
s
i
o
n
tech
n
iq
u
e
s
ca
n
b
e
class
if
i
ed
as
eith
er
lo
s
s
y
o
r
lo
s
s
les
s
.
I
n
lo
s
s
le
s
s
co
m
p
r
ess
io
n
,
th
e
o
r
ig
i
n
al
i
m
a
g
e
o
r
d
at
a
is
r
etai
n
ed
e
x
ac
tl
y
a
f
ter
d
ec
o
m
p
r
es
s
io
n
.
L
o
s
s
y
co
m
p
r
es
s
io
n
ac
h
ie
v
es b
etter
co
m
p
r
es
s
io
n
r
atio
s
w
i
th
s
o
m
e
a
m
o
u
n
t o
f
d
is
to
r
tio
n
in
t
h
e
d
ec
o
m
p
r
es
s
ed
d
ata.
T
o
g
et
th
e
b
est
c
o
m
p
r
ess
io
n
p
er
f
o
r
m
an
ce
,
co
m
p
r
es
s
in
g
C
F
A
d
ata
is
m
o
r
e
ef
f
icie
n
t
t
h
an
c
o
m
p
r
es
s
i
n
g
d
em
o
s
aick
ed
co
lo
r
i
m
ag
e
s
,
as
r
ep
r
esen
ted
in
[
4
]
.
I
n
p
ar
ticu
la
r
,
th
e
co
m
p
r
ess
io
n
s
c
h
e
m
e
i
s
m
o
r
e
e
f
f
icie
n
t
t
h
a
n
th
e
d
e
m
o
s
aic
k
i
n
g
-
f
ir
s
t
m
et
h
o
d
b
ec
au
s
e
th
e
d
e
m
o
s
aic
k
in
g
m
e
th
o
d
in
cr
ea
s
es
t
h
e
n
u
m
b
er
o
f
d
ata
p
o
in
ts
th
at
ar
e
s
o
m
e
m
ea
n
s
co
r
r
elate
d
.
I
n
th
e
ea
r
ly
lo
s
s
y
C
F
A
co
m
p
r
ess
io
n
m
et
h
o
d
s
m
e
n
tio
n
e
ab
o
v
e
[
5
]
–
[
7
]
,
th
e
R
GB
co
m
p
o
n
e
n
t
s
o
f
C
F
A
d
ata
ar
e
co
n
v
er
ted
to
d
ec
o
r
r
elate
d
R
GB
co
m
p
o
n
e
n
t
s
,
w
h
ic
h
i
s
in
d
ep
en
d
en
tl
y
e
n
co
d
ed
.
I
n
t
h
is
p
ap
er
,
w
e
p
r
o
p
o
s
e
a
n
e
w
p
r
ed
icti
v
e
co
d
in
g
s
c
h
e
m
e
d
ep
en
d
s
o
n
h
ier
ar
ch
ical
p
r
ed
ictio
n
m
et
h
o
d
an
d
a
n
e
w
co
n
tex
t a
d
a
p
tiv
e
m
et
h
o
d
w
h
ic
h
i
s
e
n
ab
led
b
y
t
h
e
h
ier
ar
ch
ical
s
c
h
e
m
e
f
o
r
R
GB
I
m
a
g
es
an
d
C
F
A
I
m
a
g
es.
Fo
r
C
F
A
I
m
ag
e
s
,
in
h
ier
ar
ch
ica
l
p
r
ed
ictio
n
,
h
alf
o
f
th
e
G
p
ix
el
s
ar
e
u
s
ed
f
o
r
th
e
p
r
ed
ictio
n
o
f
th
e
o
th
er
h
al
f
o
f
G
p
i
x
els,
b
o
th
s
ets
o
f
G
p
ix
el
s
ar
e
u
s
ed
to
p
r
ed
ict
R
p
ix
els,
an
d
Gr
ee
n
a
n
d
R
ed
p
ix
el
v
al
u
e
s
ar
e
th
en
u
s
ed
to
p
r
ed
ict
B
lu
e
p
ix
els.
Fo
r
R
GB
I
m
a
g
es,
th
e
co
m
p
o
n
en
ts
ar
e
f
ir
s
t
tr
an
s
f
o
r
m
e
d
o
r
co
n
v
er
ted
b
y
a
r
ev
e
r
s
ib
le
co
lo
r
tr
an
s
f
o
r
m
,
an
d
ea
ch
o
f
th
e
tr
a
n
s
f
o
r
m
ed
co
m
p
o
n
en
ts
ar
e
i
n
d
ep
en
d
en
tl
y
co
m
p
r
es
s
ed
b
y
th
e
ab
o
v
e
ex
p
lai
n
ed
m
eth
o
d
s
.
Du
e
to
h
i
g
h
d
e
n
s
it
y
a
n
d
h
i
g
h
l
y
c
o
r
r
elate
d
o
f
R
GB
d
ata
a
n
d
it
ca
n
n
o
t
b
e
co
m
p
r
es
s
ad
ep
tly
a
n
d
it
is
c
h
an
g
ed
to
YC
b
C
r
b
y
u
s
i
n
g
r
e
v
er
s
e
co
lo
r
tr
an
s
f
o
r
m
.
I
n
t
h
is
b
o
th
C
F
A
a
n
d
R
GB
h
ier
ar
ch
ical
s
ch
e
m
e,
alr
ea
d
y
e
n
co
d
ed
p
ix
els
ar
e
u
s
ed
f
o
r
co
n
te
x
t
ad
ap
tiv
e
m
o
d
elin
g
,
i.e
to
f
i
n
d
th
e
co
n
d
itio
n
al
p
r
o
b
ab
ilit
y
d
en
s
it
y
f
u
n
ctio
n
(
p
d
f
)
o
f
p
r
ed
ictio
n
er
r
o
r
f
o
r
th
e
g
iv
e
n
n
eig
h
b
o
r
in
g
p
ix
els.
Af
ter
t
h
e
p
r
e
d
ictio
n
an
d
co
n
te
x
t
ad
ap
ti
v
e
m
o
d
eli
n
g
,
t
h
e
p
r
e
d
ictio
n
er
r
o
r
s
also
w
i
th
t
h
e
co
n
tex
ts
ar
e
en
co
d
ed
u
s
i
n
g
a
co
n
v
e
n
tio
n
a
l
co
n
te
x
t
-
ad
ap
tiv
e
ar
it
h
m
etic
en
co
d
er
[
1
6
]
.
I
n
th
e
e
x
p
er
i
m
en
ts
,
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
is
a
n
al
y
ze
d
w
it
h
r
ec
e
n
t
p
r
ed
ictiv
e
en
co
d
in
g
m
et
h
o
d
in
[
1
0
]
,
w
h
ic
h
s
a
y
s
t
h
e
b
est
p
er
f
o
r
m
a
n
ce
a
m
o
n
g
th
e
ex
i
s
ti
n
g
m
et
h
o
d
s
,
an
d
als
o
w
it
h
th
e
m
o
d
er
n
tr
a
n
s
f
o
r
m
m
eth
o
d
in
[
8
]
.
C
o
m
p
ar
in
g
t
h
e
r
esu
l
ts
o
n
s
o
m
e
s
i
m
u
lated
C
F
A
d
ata
an
d
also
o
n
r
ea
l
C
F
A
d
ata
av
ailab
le
i
n
[
2
0
]
.
T
h
e
co
m
p
ar
is
o
n
s
h
o
w
s
th
at
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
g
iv
e
s
less
b
its
p
er
p
ix
el
o
n
all
o
f
th
e
i
m
a
g
es
r
ef
er
en
ce
d
ab
o
v
e.
T
h
e
r
em
ai
n
i
n
g
p
ar
t
o
f
th
is
p
ap
er
is
o
r
g
an
ized
as
f
o
llo
w
s
.
I
n
s
u
b
d
iv
is
io
n
2
,
th
e
s
tr
u
c
tu
r
e
o
f
o
u
r
C
F
A
e
n
co
d
er
is
p
r
esen
ted
.
T
h
en
,
Su
b
d
iv
is
io
n
3
d
ea
ls
w
it
h
p
r
ed
ictio
n
s
ch
e
m
e
an
d
Su
b
d
iv
is
io
n
4
p
r
es
en
ts
th
e
co
n
te
x
t
m
o
d
elin
g
f
o
r
ad
ap
tiv
e
en
co
d
in
g
.
E
x
p
er
i
m
e
n
ts
o
n
v
ar
io
u
s
s
i
m
u
lated
an
d
r
ea
l
C
F
A
a
n
d
R
G
B
d
ata
ar
e
ex
p
lain
ed
in
S
u
b
d
iv
is
io
n
5
,
an
d
th
e
co
n
clu
s
io
n
is
p
r
ese
n
ted
in
S
u
b
d
iv
is
io
n
6
.
2.
O
VE
RVI
E
W
O
F
P
RO
P
O
SE
D
E
NCO
D
E
R
2
.
1
.
RG
B
I
m
a
g
es
Fo
r
th
e
p
r
ed
ictio
n
o
f
a
p
ix
el
t
h
at
i
s
to
b
e
e
n
co
d
ed
,
h
er
e
we
p
r
o
p
o
s
e
a
m
et
h
o
d
w
h
ic
h
m
ak
es
u
s
e
o
f
lo
w
er
r
o
w
p
i
x
els
as
w
e
ll
as
t
h
e
u
p
p
er
an
d
also
lef
t
p
ix
el
s
F
o
r
th
e
lo
s
s
les
s
co
m
p
r
ess
io
n
o
f
co
lo
r
co
m
p
o
n
en
t
s
,
th
e
R
GB
is
f
ir
s
t
alter
ed
to
Y
b
y
a
n
R
C
T
r
an
s
f
o
r
m
an
d
Y
-
g
r
e
y
ch
a
n
n
el
i
s
d
eter
m
in
ed
b
y
a
co
n
v
en
t
io
n
al
g
r
a
y
s
ca
le
i
m
a
g
e
co
m
p
r
ess
io
n
alg
o
r
ith
m
.
T
h
e
s
ig
n
al
d
is
i
m
ila
r
it
y
is
g
e
n
er
all
y
m
u
c
h
s
m
a
ll
er
th
an
t
h
at
o
f
R
GB
s
ig
n
al,
b
u
t
s
t
ill
b
ig
g
er
n
ea
r
t
h
e
ed
g
es
i
n
th
e
ca
s
e
o
f
ch
r
o
m
i
n
an
ce
c
h
an
n
el
s
(
)
.
Fo
r
m
o
r
e
d
ef
in
ed
p
r
ed
ictio
n
o
f
th
ese
s
i
g
n
als,
a
n
d
also
f
o
r
h
ier
ar
ch
ical
m
o
d
e
lin
g
o
f
p
r
ed
ictio
n
er
r
o
r
s
,
w
e
u
s
e
t
h
e
h
ier
ar
ch
ica
l
s
ch
e
m
e,
th
e
c
h
r
o
m
a
i
m
a
g
e
is
s
p
litt
ed
in
to
t
w
o
s
u
b
i
m
a
g
es
;
i.e
.
a
p
air
o
f
ev
en
n
u
m
b
er
ed
r
o
w
s
an
d
a
p
air
o
f
o
d
d
n
u
m
b
er
ed
r
o
w
s
r
es
p
ec
tiv
el
y
.
T
h
e
ev
en
r
o
w
s
u
b
i
m
a
g
e
is
p
r
e
-
ar
r
an
g
ed
,
th
e
n
w
e
ca
n
m
a
k
e
u
s
e
o
f
all
th
e
p
ix
els i
n
f
o
r
th
e
p
r
ed
ictio
n
o
f
a
p
ix
els i
n
th
e
o
d
d
r
o
w
s
u
b
i
m
a
g
e
.
2
.
2
.
CF
A
I
m
a
g
es
T
h
e
p
r
o
p
o
s
ed
en
co
d
er
’
s
s
tr
u
ctu
r
e
is
s
h
o
w
n
i
n
Fi
g
u
r
e
2
,
w
h
ic
h
h
as
h
ier
ar
ch
ical
p
r
ed
icto
r
,
a
co
n
v
e
n
tio
n
al
g
r
a
y
s
ca
le
e
n
co
d
er
,
an
d
a
co
n
tex
t
-
ad
ap
tiv
e
ar
it
h
m
e
tic
en
co
d
er
.
T
h
e
in
p
u
t
C
F
A
i
m
a
g
e
w
h
ich
i
s
in
Fig
u
r
e
1
,
th
e
G
p
ix
els
in
th
e
o
d
d
r
o
w
s
(
G
1
p
ix
els)
ar
e
en
c
o
d
ed
f
ir
s
t
u
s
in
g
a
co
n
v
e
n
tio
n
al
g
r
a
y
s
ca
le
co
d
er
m
et
h
o
d
.
Nex
t,
t
h
e
y
ar
e
u
s
ed
f
o
r
th
e
p
r
ed
ictio
n
o
f
t
h
e
G
p
i
x
els
i
n
t
h
e
e
v
en
n
u
m
b
er
ed
r
o
w
s
(
G
2
p
ix
el
s
)
,
w
h
ich
p
r
o
d
u
ce
s
th
e
p
r
ed
ictio
n
er
r
o
r
t
h
at
i
s
r
ep
r
esen
ted
a
s
eG
2.
T
h
i
r
d
,
th
e
in
ter
p
o
latin
g
t
h
e
G
p
i
x
els,
to
f
ill
i
n
th
e
G
v
alu
e
s
at
th
e
p
o
s
itio
n
s
o
f
th
e
R
ed
an
d
B
lu
e
p
ix
els.
Fo
u
r
t
h
,
th
e
p
r
ev
io
u
s
l
y
i
n
ter
p
o
lated
G
p
ix
els
ar
e
s
u
b
tr
ac
ted
f
r
o
m
t
h
e
R
a
n
d
B
p
ix
els,
p
r
o
d
u
cin
g
t
h
e
d
if
f
er
en
ce
ΔR
a
n
d
ΔB
.
I
t
is
to
n
o
te
th
at
ΔR
a
n
d
ΔB
ar
e
u
s
ed
as
a
n
alter
n
ati
v
e
o
f
R
a
n
d
B
,
r
esp
ec
tiv
el
y
,
to
ex
p
lo
it th
e
c
h
an
n
el
s
co
r
r
elatio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
4
,
No
.
1
,
Ma
r
ch
201
5
:
24
–
31
26
Fig
u
r
e
2.
T
h
e
p
r
o
p
o
s
ed
en
co
d
er
’
s
s
tr
u
ct
u
r
e
Fif
t
h
,
t
h
e
ΔR
v
al
u
es
ar
e
p
r
ed
icted
f
r
o
m
t
h
e
p
r
ev
io
u
s
l
y
e
n
c
o
d
ed
n
eig
h
b
o
r
in
g
ΔR
v
al
u
es,
alo
n
g
w
it
h
th
e
in
f
o
r
m
at
io
n
o
b
tain
ed
f
r
o
m
th
e
G
p
ix
els,
an
d
th
e
p
r
ed
ictio
n
er
r
o
r
w
h
ic
h
is
r
ep
r
esen
ted
as
eΔ
R
is
o
b
tain
ed
.
A
tla
s
t,
all
o
f
t
h
e
p
r
ev
io
u
s
l
y
e
n
co
d
ed
p
ix
els
(
w
h
ich
in
cl
u
d
e
s
th
e
G,
ΔR
,
an
d
p
r
ec
ed
in
g
Δ
B
p
ix
els)
ar
e
u
s
ed
to
s
o
lv
e
a
n
ap
p
r
o
p
r
iate
p
r
ed
ict
o
r
f
o
r
a
g
i
v
e
n
ΔB
,
a
n
d
t
h
e
p
r
ed
i
ctio
n
er
r
o
r
eΔ
B
is
g
e
n
er
ated
.
T
h
e
p
r
ed
ictio
n
b
lo
c
k
o
b
tain
s
th
e
er
r
o
r
s
ig
n
als,
i.e
.
,
eG
2
,
eΔ
R
,
an
d
eΔ
B
a
r
e
f
ed
in
to
th
e
co
n
te
x
t
-
ad
ap
tiv
e
ar
ith
m
et
i
c
en
co
d
er
.
Alg
o
rit
h
m
(
a)
: T
o
f
in
d
th
e
d
ir
ec
tio
n
o
f
i,
j
.
if
(
i,
j
)
-
̂
|+
< |
(
i,j
)
-
̂
|
t
h
en
d
ir
(
i,
j
)
H
else
d
ir
(
i,
j
)
V
en
d
if
Nea
r
th
e
ed
g
es
th
e
lar
g
e
p
r
ed
ictio
n
er
r
o
r
s
ar
e
ex
p
ec
ted
a
n
d
w
i
th
i
n
tex
tu
r
ed
ar
ea
s
ev
e
n
if
a
m
u
c
h
m
o
r
e
e
lab
o
r
ate
p
r
ed
icto
r
is
u
s
ed
,
w
h
ich
s
tr
ictl
y
d
e
g
r
ad
es
th
e
p
er
f
o
r
m
an
ce
o
f
a
co
n
v
e
n
tio
n
al
e
n
tr
o
p
y
m
eth
o
d
.
Ho
w
e
v
er
,
if
w
e
ab
le
to
esti
m
ate
th
e
p
d
f
o
f
t
h
e
er
r
o
r
f
o
r
th
e
g
i
v
en
n
ei
g
h
b
o
r
p
ix
el
s
,
m
o
r
e
ef
f
icie
n
t
en
co
d
i
n
g
m
et
h
o
d
is
p
o
s
s
ib
le
as
m
en
tio
n
ed
in
[
1
8
]
.
Mo
r
e
p
r
ec
is
el
y
,
w
h
e
n
w
e
e
n
co
d
e
an
p
r
ed
ictio
n
er
r
o
r
at
a
p
ix
el
p
o
s
itio
n
n
,
w
e
ca
n
m
a
k
e
u
s
e
o
f
th
e
i
n
f
o
r
m
atio
n
f
r
o
m
t
h
e
al
r
ea
d
y
e
n
co
d
ed
n
eig
h
b
o
r
in
g
p
i
x
els
a
s
th
e
co
n
tex
t
Cn
.
I
n
n
e
w
w
o
r
d
s
,
w
e
ca
n
p
u
t
u
p
th
e
p
d
f
P
(
en
|
Cn
)
w
h
ile
e
n
c
o
d
in
g
,
w
h
ic
h
is
u
s
ed
f
o
r
t
h
e
a
d
ap
tiv
e
ar
ith
m
etic
co
d
in
g
.
Hen
ce
,
i
f
th
e
co
n
te
x
t
m
o
d
eli
n
g
is
p
er
f
ec
t
a
n
d
co
r
r
e
ct,
ev
en
lar
g
e
er
r
o
r
s
ar
e
ex
p
ec
ted
to
b
e
w
it
h
th
e
lo
w
e
n
tr
o
p
y
a
n
d
th
u
s
th
e
co
n
t
ex
t a
d
ap
tiv
e
e
n
co
d
er
p
r
o
d
u
ce
s
less
b
it
s
t
h
an
a
n
o
r
m
al
e
n
tr
o
p
y
co
d
er
.
T
h
e
d
etails
o
f
co
n
tex
t
m
o
d
eli
n
g
f
o
r
th
e
C
F
A
d
ata
w
i
ll b
e
d
escr
ib
ed
in
s
u
b
d
iv
is
io
n
4
.
3.
H
I
E
RAR
CH
I
CA
L
P
RE
DIC
T
I
O
N
O
F
CF
A
AND
R
G
B
DATA
3
.
1
.
P
re
dict
io
n f
o
r
CF
A
Da
t
a
As
s
h
o
w
n
i
n
Fi
g
u
r
e
1
,
w
e
d
e
n
o
te
a
m
o
s
aic
i
m
ag
e
a
s
co
n
s
i
s
ts
o
f
f
o
u
r
s
u
b
i
m
a
g
es
:
G
1
,
G
2
,
R
,
an
d
B
.
As
m
e
n
tio
n
ed
p
r
ev
io
u
s
l
y
,
w
e
f
ir
s
t
en
co
d
e
G
1
p
ar
t
b
y
a
g
r
ay
s
ca
le
co
m
p
r
es
s
io
n
m
eth
o
d
J
P
E
G
-
L
o
s
s
les
s
,
an
d
th
en
h
i
er
ar
ch
ica
l sc
h
e
m
e
o
f
p
r
ed
ictio
n
o
f
o
th
er
s
u
b
i
m
a
g
es
f
o
llo
w
s
.
Fig
u
r
e
3
.
I
llu
s
tr
atio
n
f
o
r
th
e
e
x
p
lan
atio
n
o
f
p
r
ed
ictio
n
o
f
G
2
p
ix
el
f
r
o
m
t
h
e
n
e
ig
h
b
o
r
in
g
p
ix
els.
G
1
p
ix
el
s
ar
e
d
en
o
ted
as lig
h
t g
r
ee
n
,
an
d
th
e
G
2
as d
ar
k
g
r
ee
n
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
A
n
E
n
h
a
n
ce
d
Lo
s
s
less
C
o
lo
r
F
ilter
A
r
r
a
y
I
ma
g
e
C
o
mp
r
ess
i
o
n
B
a
s
ed
o
n
P
r
ed
i
ctive
…
(
L.
M.
V
a
r
a
la
ksh
mi
)
27
3
.
1
.
1
.
P
re
dict
io
n o
f
G
2
L
et
u
s
r
ef
lec
t o
n
th
e
s
it
u
atio
n
t
h
at
a
p
ix
el
d
en
o
ted
as
x
in
F
ig
u
r
e
3
is
en
co
d
ed
.
A
lg
o
rit
h
m
(
b)
T
o
ca
lcu
late
th
e
o
v
er
all
p
ix
el
p
r
ed
ictio
n
if
d
ir
(
i
-
1
,
j
)
=
H
o
r
d
ir
(
i,
j
-
1
)
=
H
th
e
n
C
alcu
late
d
ir
(
i,j
)
b
y
A
l
g
o
r
ith
m
(
a)
Pro
g
r
am
d
ir
(
i,j
)
if
d
ir
(
i,
j
)
=
H
th
en
̂
̂
else
̂
̂
en
d
if
else
̂
̂
C
alcu
late
d
ir
(
i,j
)
b
y
A
l
g
o
r
ith
m
(
a)
en
d
if
No
te
th
at
li
g
h
t
g
r
ee
n
b
o
x
es
a
r
e
th
e
p
ix
els
d
e
n
o
ted
in
t
h
e
o
d
d
r
o
w
s
(
G
1
)
an
d
d
ar
k
g
r
ee
n
b
o
x
es
ar
e
r
ep
r
esen
ted
in
th
e
e
v
e
n
r
o
w
s
(
G
2
)
.
T
o
p
r
ed
ict
x
,
w
e
ca
n
u
s
e
th
e
alr
ea
d
y
en
co
d
ed
p
ix
els
o
f
G
2
s
u
ch
a
s
n
w
,
n,
n
e,
w
,
a
n
d
all
t
h
e
n
e
ig
h
b
o
r
in
g
p
ix
els
i
n
G
1
.
A
s
i
n
co
n
v
en
tio
n
al
p
r
ed
ictio
n
p
r
o
ce
s
s
,
w
e
d
e
f
in
e
f
o
u
r
v
ar
itie
s
o
f
d
ir
ec
tio
n
al
p
r
ed
icto
r
s
(
h
o
r
izo
n
t
al
d
iag
o
n
al,
v
er
tica
l,
r
ig
h
t d
iag
o
n
al,
le
f
t d
iag
o
n
al)
as in
E
q
u
atio
n
(
1
)
.
{
}
{
}
(
1
)
W
h
ich
co
n
s
i
s
t
o
f
t
h
e
ad
j
ac
en
t
p
ix
els
to
t
h
e
x
,
in
to
t
h
e
r
esp
ec
tiv
e
d
ir
ec
tio
n
s
.
I
n
t
h
is
p
r
o
p
o
s
ed
s
ch
e
m
e,
it
is
n
o
t
n
ee
d
ed
to
ch
o
o
s
e
o
n
e
o
u
t o
f
th
ese
f
o
u
r
p
r
ed
icto
r
s
,
b
u
t
w
e
.
T
o
r
ea
lize
th
is
d
esig
n
,
w
e
d
ef
i
n
e
a
v
ar
iab
le
f
o
r
th
e
c
h
o
o
s
e
an
y
t
w
o
b
est
p
r
ed
icto
r
s
an
d
co
m
b
in
e
t
h
e
m
w
it
h
ap
p
r
o
p
r
iate
w
ei
g
h
ts
.
T
h
e
ch
o
ice
o
f
p
r
ed
icto
r
s
an
d
w
ei
g
h
ts
i
s
o
f
co
u
r
s
e
d
ep
en
d
s
o
n
th
e
d
ir
ec
tio
n
o
f
ed
g
e
s
ar
o
u
n
d
th
e
x
.
E
d
g
e
d
ir
ec
tiv
it
y
ar
o
u
n
d
x
as r
ep
r
esen
ted
as
:
=
W
h
er
e
a
=
h
,
v
,
dr
,
o
r
dl
.
3
.
1
.
2
.
I
nte
rpo
la
t
io
n o
f
G
re
en
Va
lu
es in P
o
s
it
io
ns
o
f
R
a
nd
B
As
m
e
n
tio
n
ed
ab
o
v
e,
th
e
b
lu
e
an
d
r
ed
p
ix
els
ar
e
n
o
t
e
n
co
d
ed
d
ir
ec
tly
,
b
u
t
ΔR
=
R
–
̂
an
d
ΔB
=
B
−
̂
ar
e
en
co
d
ed
d
ir
ec
tly
an
d
als
o
p
r
ed
icted
,
w
h
er
e
̂
is
a
g
r
ee
n
in
ter
p
o
latio
n
v
al
u
e.
B
ec
au
s
e,
all
th
e
clo
s
es
t
f
o
u
r
n
ei
g
h
b
o
r
s
ar
e
av
ailab
le
i
n
th
i
s
ca
s
e,
w
e
i
n
ter
p
o
late
(
p
r
ed
ict)
o
n
l
y
in
to
h
o
r
izo
n
tal
an
d
v
er
tical
d
ir
ec
tio
n
s
d
ef
in
ed
as i
n
E
q
u
atio
n
(
2
)
.
{
}
{
}
(
2
)
3
.
1
.
3
.
Pr
edict
io
n o
f
Red
P
ix
els a
nd
B
lue P
ix
els
On
ce
o
b
tain
i
n
g
t
h
e
G
v
al
u
es
in
t
h
e
p
o
s
itio
n
s
o
f
th
e
R
a
n
d
B
,
w
e
ca
n
ca
lcu
late
t
h
e
ΔR
a
n
d
ΔB
.
Fo
r
p
r
ed
ictin
g
th
e
ΔR
i
n
th
e
p
o
s
iti
o
n
m
ar
k
ed
as
x
,
th
e
p
r
ed
icto
r
s
ar
e
s
i
m
p
l
y
t
h
e
n
ei
g
h
b
o
r
in
g
ΔR
s
in
t
h
e
m
atc
h
i
n
g
d
ir
ec
tio
n
s
as i
n
E
q
u
atio
n
(
3
)
.
{
}
{
}
(
3
)
3
.
2
.
P
re
dict
io
n M
et
ho
d f
o
r
RG
B
I
m
a
g
e
T
h
e
ch
r
o
m
in
a
n
ce
(
i.e
)
ch
r
o
m
a
ch
a
n
n
els
C
u
an
d
C
v
w
h
ich
r
es
u
lts
f
r
o
m
t
h
e
R
C
T
u
s
u
all
y
h
av
e
d
if
f
er
e
n
t
lab
els
f
r
o
m
Y
(
l
u
m
a)
,
an
d
also
d
if
f
er
en
t
f
r
o
m
th
e
o
r
ig
i
n
al
co
lo
r
p
lan
es
R
,
G
,
an
d
B
in
th
i
s
h
ier
ar
ch
ical
d
ec
o
m
p
o
s
itio
n
s
c
h
e
m
e.
T
h
e
o
v
er
all
s
i
g
n
al
d
is
s
i
m
ilar
it
y
is
m
as
k
ed
b
y
t
h
e
co
lo
r
tr
an
s
f
o
r
m
i
n
th
e
ch
r
o
m
i
n
an
ce
c
h
a
n
n
el
s
,
b
u
t
t
h
e
d
is
p
ar
ity
i
s
s
ti
ll
lar
g
e
n
ea
r
th
e
o
b
j
ec
t
b
o
u
n
d
ar
ies.
As
a
r
esu
lt,
t
h
e
p
r
ed
ictio
n
er
r
o
r
s
in
a
c
h
r
o
m
a
c
h
a
n
n
el
ar
e
m
u
ch
r
ed
u
ce
d
i
n
a
s
m
o
o
t
h
(
i:
e)
h
o
r
izo
n
tal
r
e
g
io
n
,
b
u
t
r
e
m
a
in
r
ea
s
o
n
ab
l
y
g
r
ea
t
n
ea
r
th
e
ed
g
e
o
r
w
ith
in
a
te
x
t
u
r
e
r
eg
io
n
ex
p
lai
n
ed
in
A
l
g
o
r
it
h
m
(
a)
an
d
(
b
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
4
,
No
.
1
,
Ma
r
ch
201
5
:
24
–
31
28
T
h
e
p
d
f
o
f
p
r
ed
ictio
n
er
r
o
r
f
o
r
b
etter
c
o
n
tex
t
m
o
d
eli
n
g
,
alo
n
g
w
it
h
th
e
p
r
ec
is
e
p
r
ed
ictio
n
is
esti
m
ated
f
o
r
t
h
e
e
f
f
icie
n
t
co
m
p
r
e
s
s
io
n
.
Her
e,
w
e
ad
v
is
e
(
p
r
o
p
o
s
e)
a
h
ier
ar
ch
ical
d
ec
o
m
p
o
s
itio
n
s
ch
e
m
e
th
at
is
p
ix
els
in
an
k
e
y
(
i:e)
in
p
u
t
im
a
g
e
X
is
s
p
litt
in
g
in
to
t
wo
s
u
b
i
m
a
g
es:
an
ev
e
n
s
u
b
i
m
ag
e
an
d
an
o
d
d
s
u
b
i
m
ag
e
.
A
n
e
v
e
n
s
u
b
i
m
a
g
e
is
en
co
d
ed
f
ir
s
t a
n
d
i
s
u
s
ed
to
p
r
ed
ict
th
e
p
ix
els in
o
d
d
s
u
b
i
m
a
g
e
.
I
n
ad
d
itio
n
,
is
also
u
s
ed
to
f
i
n
d
th
e
f
i
g
u
r
es
o
f
p
r
ed
ictio
n
er
r
o
r
s
o
f
.
Fo
r
th
e
co
m
p
r
ess
i
o
n
o
f
p
ix
els
u
s
i
n
g
,
d
ir
ec
tio
n
al
p
r
ed
i
ctio
n
is
w
o
r
k
ed
to
s
ta
y
a
w
a
y
f
r
o
m
t
h
e
lar
g
e
p
r
ed
ictio
n
er
r
o
r
s
clo
s
e
to
th
e
ed
g
es.
Fo
r
ea
ch
p
i
x
el,
t
h
e
h
o
r
izo
n
tal
p
r
ed
icto
r
̂
(
i,
j
)
a
n
d
v
er
tical
p
r
ed
icto
r
̂
(
i,
j
)
ar
e
d
ef
in
ed
a
s
i
n
E
q
u
atio
n
(
4
)
.
̂
)
,
(
4
)
A
n
d
o
n
e
a
m
o
n
g
th
e
m
is
s
elec
t
ed
to
p
r
ed
ict
(
i,j)
.
T
h
e
m
o
s
t i
m
p
o
r
ta
n
t o
n
e
is
t
h
e
h
o
r
izo
n
tal
p
r
ed
icto
r
b
ec
au
s
e
it
w
ill
b
e
m
o
r
e
ac
cu
r
ate
o
n
l
y
w
h
en
th
er
e
i
s
a
s
tr
o
n
g
h
o
r
izo
n
tal
ed
g
e
s
.
T
o
r
ea
lize
th
is
d
esig
n
,
w
e
d
ef
i
n
e
a
v
ar
iab
le
f
o
r
t
h
e
d
ir
ec
tio
n
o
f
ed
g
e
at
ea
c
h
p
i
x
el
d
ir
(
i,
j
)
,
w
h
ic
h
i
s
g
iv
e
n
e
ith
er
Ho
r
izo
n
tal
o
r
Ver
tical
.
T
o
d
ec
id
e
th
e
d
ir
ec
tio
n
o
f
i,
j
an
d
it
is
ex
p
lai
n
ed
in
A
l
g
o
r
ith
m
(
a
)
.
4.
SI
M
UL
AT
I
O
N
R
E
S
UL
T
S
4
.
1
.
F
o
r
RG
B
L
m
a
g
es by
RCT u
s
ing
Co
nte
x
t
Co
din
g
On
t
h
e
v
ar
io
u
s
te
s
t
i
m
ag
e
s
,
th
is
a
lg
o
r
it
h
m
is
ap
p
lied
,
w
h
ich
is
w
id
el
y
u
s
ed
f
o
r
t
h
e
lo
s
s
les
s
co
m
p
r
es
s
io
n
.
I
n
all
th
e
s
i
m
u
la
tio
n
r
es
u
lts
,
th
e
p
ar
a
m
eter
T
1
in
A
l
g
o
r
ith
m
(
b
)
an
d
t
h
e
q
u
an
tit
y
o
f
co
n
te
x
t
s
ar
e
ass
i
g
n
ed
to
3
a
n
d
5
.
T
h
e
co
n
d
itio
n
al
p
r
o
b
ab
ilit
y
d
en
s
it
y
f
u
n
ct
io
n
h
a
s
b
ee
n
s
i
m
u
lated
as
m
e
n
tio
n
ed
in
Fi
g
u
r
e
4
.
T
h
e
class
ic
i
m
a
g
es
w
h
ich
h
as
a
co
m
b
i
n
atio
n
o
f
B
lu
e,
Gr
ee
n
an
d
R
ed
an
d
it
is
d
en
o
ted
(
ch
an
g
ed
)
in
to
Y
th
en
co
m
p
r
e
s
s
io
n
is
p
er
f
o
r
m
ed
an
d
ar
ith
m
etic
co
d
in
g
h
a
s
b
ee
n
ap
p
lied
to
m
ea
s
u
r
e/ca
l
cu
late
b
it
r
ate
an
d
P
SNR
.
(
a)
I
n
p
u
t I
m
ag
e
(
b
)
C
o
n
tex
t
(
c)
C
o
n
d
itio
n
al
P
d
f
Fig
u
r
e
4
.
A
n
e
x
a
m
p
le
o
f
lo
ca
l
ac
tiv
it
y
(
co
n
te
x
t)
an
d
p
r
o
b
ab
i
lit
y
o
f
er
r
o
r
d
ep
en
d
in
g
o
n
co
n
t
ex
t
T
h
e
r
esu
lts
w
h
ic
h
ar
e
s
i
m
u
late
d
ar
e
s
u
m
m
ar
ized
in
T
ab
le
1,
T
ab
le
2
an
d
T
ab
le
3
w
h
ich
co
m
p
ar
e
s
th
e
co
m
p
r
es
s
ed
b
it r
ates,
P
SNR
an
d
C
o
d
in
g
ti
m
e
w
it
h
ex
i
s
ti
n
g
m
et
h
o
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
A
n
E
n
h
a
n
ce
d
Lo
s
s
less
C
o
lo
r
F
ilter
A
r
r
a
y
I
ma
g
e
C
o
mp
r
ess
i
o
n
B
a
s
ed
o
n
P
r
ed
i
ctive
…
(
L.
M.
V
a
r
a
la
ksh
mi
)
29
T
ab
le
1
.
C
o
m
p
ar
is
o
n
o
f
C
o
m
p
r
ess
ed
B
it R
ates (
B
p
p
)
Fo
r
C
l
ass
ic
I
m
a
g
es
S
i
z
e
C
A
L
IC
J
PE
G X
R
Pr
o
p
o
sed
L
e
n
a
5
1
2
×
5
1
2
1
3
.
1
7
8
7
1
4
.
0
9
4
2
1
3
.
5
1
6
2
P
e
p
p
e
r
s
5
1
2
×
5
1
2
1
3
.
8
6
6
1
1
5
.
3
2
4
5
0
9
.
0
7
0
3
Ey
e
5
1
2
×
5
1
2
1
8
.
1
5
1
1
1
8
.
2
5
5
3
1
0
.
0
6
4
1
S
t
r
a
w
b
e
r
r
y
5
1
2
×
5
1
2
1
4
.
9
5
6
7
1
5
.
1
4
0
8
1
2
.
8
9
2
7
T
ab
le
2
.
P
ea
k
Sig
n
a
l to
No
is
e
R
atio
(
d
B
)
Fo
r
C
lass
ic
I
m
ag
e
s
S
i
z
e
C
A
L
IC
Pr
o
p
o
sed
L
e
n
a
5
1
2
×
5
1
2
1
1
.
4
5
8
9
1
3
.
5
5
0
0
P
e
p
p
e
r
s
5
1
2
×
5
1
2
0
7
.
7
3
1
2
1
0
.
1
2
0
9
Ey
e
5
1
2
×
5
1
2
0
7
.
1
9
6
2
1
2
.
0
5
3
2
S
t
r
a
w
b
e
r
r
y
5
1
2
×
5
1
2
0
6
.
6
0
5
7
1
1
.
8
9
2
7
T
ab
le
3
.
C
o
m
p
ar
is
o
n
o
f
C
P
U
T
im
es
(
Seco
n
d
s
)
On
A
P
c
I
n
te
l Co
r
e
-
I3
-
2
.
2
0
Gh
z
C
p
u
I
n
p
u
t
I
mag
e
s
En
t
r
o
p
y
o
f
R
e
d
C
h
a
n
n
e
l
En
t
r
o
p
y
o
f
G
r
e
e
n
C
h
a
n
n
e
l
En
t
r
o
p
y
o
f
B
l
u
e
C
h
a
n
n
e
l
Jo
i
n
t
E
n
t
r
o
p
y
1
7
.
5
5
3
5
7
.
6
1
2
7
.
6
4
0
6
1
5
.
5
8
4
2
7
.
6
2
8
7
.
6
6
1
7
.
5
7
8
9
1
5
.
4
4
9
3
7
.
0
4
8
7
7
.
2
9
6
2
7
.
1
5
6
1
1
2
.
1
5
8
4
7
.
4
2
6
4
7
.
4
5
6
8
7
.
3
4
0
6
1
4
.
5
8
5
4
.
1
.
1
.
Co
m
pres
s
ed
I
m
a
g
es
Fig
u
r
e
5
.
T
h
e
C
lass
ic
i
m
ag
e
s
s
et
1
4
.
2
.
F
o
r
Co
lo
r
F
ilte
r
Arr
a
y
I
m
a
g
es
Fig
u
r
e
6
.
C
r
o
p
p
ed
p
ar
t o
f
C
FA
i
m
a
g
e
Fig
u
r
e
7.
Sp
litt
in
g
o
f
c
h
an
n
el
s
a
)
C
alcu
latio
n
o
f
C
o
m
p
r
ess
ed
b
it r
ates
T
ab
le
4
s
h
o
w
s
t
h
e
co
m
p
r
es
s
ed
b
it
r
ate
f
o
r
ea
ch
co
lo
r
f
ilter
ar
r
ay
i
m
ag
e
w
h
er
e
b
p
p
r
ep
r
esen
t
s
t
h
e
n
u
m
b
er
o
f
b
its
o
f
i
n
f
o
r
m
atio
n
s
to
r
ed
p
er
p
ix
el
o
f
an
i
m
a
g
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
4
,
No
.
1
,
Ma
r
ch
201
5
:
24
–
31
30
T
ab
le
4
.
C
alcu
latio
n
o
f
C
o
m
p
r
ess
ed
b
it r
ates
I
mag
e
s
S
i
z
e
JP
EG
-
XR
P
r
o
p
o
se
d
1
7
6
8
×
5
1
2
8
.
8
3
3
1
7
.
6
4
6
2
2
7
6
8
×
5
1
2
8
.
8
2
9
6
7
.
6
8
0
6
3
7
6
8
×
5
1
2
8
.
1
6
9
8
7
.
3
3
8
5
4
7
6
8
×
5
1
2
7
.
9
2
6
3
7
.
5
6
2
3
A
v
g
.
8
.
4
3
9
7
7
.
5
5
7
1
b
)
C
alcu
latio
n
o
f
C
o
m
p
r
es
s
io
n
R
atio
T
ab
le
5
s
h
o
w
s
t
h
e
co
m
p
r
es
s
i
o
n
r
atio
,
it is
r
atio
b
et
w
ee
n
t
h
e
s
ize
o
f
an
o
r
ig
i
n
al
i
m
a
g
e
to
th
e
s
ize
o
f
th
e
co
m
p
r
ess
ed
i
m
a
g
e
.
T
ab
le
5
.
C
o
m
p
r
ess
io
n
R
at
io
I
mag
e
s
S
i
z
e
C
o
mp
r
e
ssi
o
n
R
a
t
i
o
1
7
6
8
×
5
1
2
1
.
9
6
7
5
2
7
6
8
×
5
1
2
1
.
5
8
4
1
3
7
6
8
×
5
1
2
1
.
3
7
9
6
4
7
6
8
×
5
1
2
1
.
8
6
1
3
c
)
E
n
tr
o
p
y
o
f
C
h
a
n
n
els
T
ab
le
6
.
E
n
tr
o
p
y
o
f
C
h
an
n
el
s
I
mag
e
s
En
t
r
o
p
y
o
f
R
e
d
c
h
a
n
n
e
l
En
t
r
o
p
y
o
f
B
l
u
e
C
h
a
n
n
e
l
En
t
r
o
p
y
o
f
G
r
e
e
n
C
h
a
n
n
e
l
Jo
i
n
t
E
n
t
r
o
p
y
1
7
.
5
5
3
5
7
.
6
4
0
6
7
.
6
1
2
0
1
5
.
5
8
4
2
7
.
6
2
8
0
7
.
5
7
8
9
7
.
6
6
1
0
1
5
.
4
4
9
3
7
.
0
4
8
7
7
.
1
5
6
1
7
.
2
9
6
2
1
2
.
1
5
8
4
7
.
4
2
9
4
7
.
3
4
0
6
7
.
4
5
6
8
1
4
.
5
8
5
d
)
Me
asu
r
e
m
e
n
t o
f
P
SNR
a
n
d
MSE
T
ab
le
7
r
ep
r
esen
ts
t
h
e
m
ea
s
u
r
e
m
en
t
o
f
Me
an
s
q
u
ar
e
er
r
o
r
an
d
p
ea
k
s
i
g
n
a
l
to
n
o
is
e
r
atio
.
I
t
s
h
o
w
s
th
at
h
ig
h
er
th
e
P
SN
R
,
lo
w
er
will b
e
m
ea
n
s
q
u
ar
e
er
r
o
r
.
T
ab
le
7
.
Me
asu
r
e
m
en
t o
f
P
SNR
an
d
MSE
I
mag
e
s
S
i
z
e
JP
EG
-
XR
P
r
o
p
o
se
d
M
S
E
P
S
N
R
M
S
E
P
S
N
R
1
7
6
8
×
5
1
2
1
.
7
37
1
.
1
40
2
7
6
8
×
5
1
2
1
.
8
36
1
.
6
38
3
7
6
8
×
5
1
2
1
.
4
35
1
.
3
37
4
7
6
8
×
5
1
2
2
.
2
33
1
.
8
35
5
.
CO
NCLUS
I
O
N
A
n
e
w
lo
s
s
le
s
s
co
m
p
r
es
s
io
n
a
lg
o
r
ith
m
f
o
r
th
e
B
a
y
er
-
p
atter
n
ed
C
F
A
i
m
a
g
es
h
as
b
ee
n
p
r
o
p
o
s
ed
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
p
r
ed
icts
th
e
co
lo
r
co
m
p
o
n
en
t
s
in
h
ier
ar
ch
i
ca
l
m
an
n
er
b
ased
o
n
co
n
tex
t
ad
ap
tiv
e
ar
ith
m
e
tic
co
d
in
g
.
I
n
h
ier
ar
ch
ical
p
r
ed
ictio
n
,
w
e
e
n
co
d
e
th
e
h
al
f
o
f
th
e
g
r
ee
n
s
a
m
p
les
u
s
i
n
g
a
co
n
v
en
tio
n
al
g
r
a
y
s
ca
le
en
co
d
er
,
an
d
o
th
er
h
al
f
o
f
t
h
e
g
r
ee
n
s
a
m
p
les
ar
e
en
co
d
ed
b
ased
o
n
f
ir
s
t
h
al
f
o
f
t
h
e
e
n
co
d
ed
g
r
ee
n
s
a
m
p
le
s
.
R
ed
p
ix
els
ar
e
p
r
ed
ict
ed
u
s
i
n
g
th
e
g
r
ee
n
s
a
m
p
les
a
n
d
b
lu
e
p
ix
els
ar
e
p
r
ed
icted
u
s
i
n
g
r
ed
an
d
g
r
ee
n
s
a
m
p
le
s
.
Fo
r
r
ed
u
cin
g
th
e
p
r
ed
ictio
n
r
esid
u
al.
E
d
g
e
d
ir
ec
ti
v
it
y
is
co
n
s
id
er
ed
.
T
h
e
p
r
o
p
o
s
ed
s
ch
e
m
e
is
te
s
ted
o
n
s
i
m
u
lated
d
ata
th
at
h
as
b
ee
n
w
id
el
y
u
s
ed
an
d
o
n
th
e
r
ea
l
C
F
A
i
m
ag
e
s
an
d
s
o
m
e
ad
d
iti
o
n
al
h
i
g
h
-
r
eso
l
u
tio
n
s
i
m
u
lated
C
F
A
i
m
a
g
es.
T
h
e
r
esu
lt
s
s
h
o
w
t
h
at
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
p
r
o
v
id
es
r
ed
u
ce
d
b
it
s
p
er
p
ix
el
t
h
an
t
h
e
tr
an
s
f
o
r
m
-
b
ased
m
et
h
o
d
an
d
o
th
er
ex
i
s
ti
n
g
m
et
h
o
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
A
n
E
n
h
a
n
ce
d
Lo
s
s
less
C
o
lo
r
F
ilter
A
r
r
a
y
I
ma
g
e
C
o
mp
r
ess
i
o
n
B
a
s
ed
o
n
P
r
ed
i
ctive
…
(
L.
M.
V
a
r
a
la
ksh
mi
)
31
RE
F
E
R
E
NC
E
S
[1
]
W
.
J.
Ya
n
g
,
e
t
a
l.
,
“
Un
iv
e
rsa
l
c
h
ro
m
a
su
b
sa
m
p
li
n
g
stra
teg
y
f
o
r
c
o
m
p
re
ss
in
g
m
o
sa
ic
v
id
e
o
se
q
u
e
n
c
e
s
w
it
h
a
rb
it
ra
ry
RG
B
c
o
lo
r
f
il
ter
a
rra
y
s
in
H.2
6
4
/A
V
C
”
,
IEE
E
T
ra
n
s.
Circ
u
i
ts
S
y
st.
Vi
d
e
o
T
e
c
h
n
o
l
.
,
V
o
l
.
2
3
,
N
o
.
4
,
p
p
.
5
9
1
-
6
0
6
,
A
p
r
2
0
1
3
.
[2
]
D.
L
e
e
a
n
d
K.
N.
P
lata
n
i
o
ti
s,
“
L
o
ss
les
s
c
o
m
p
re
ss
io
n
o
f
HD
R
c
o
lo
r
f
il
ter
a
rra
y
i
m
a
g
e
f
o
r
th
e
d
ig
it
a
l
c
a
m
e
r
a
p
ip
e
li
n
e
”
,
S
i
g
n
a
l
Pr
o
c
e
ss
in
g
:
Ima
g
e
Co
mm
u
n
ica
ti
o
n
,
V
o
l
.
2
7
,
N
o
.
6
,
p
p
.
6
3
7
-
6
4
9
,
Ju
l
2
0
1
2
.
[3
]
H.
S
.
M
a
lv
a
r
a
n
d
G
.
J.
S
u
ll
i
v
a
n
,
“
Pro
g
re
ss
ive
-
to
-
lo
ss
les
s
c
o
mp
re
ss
io
n
o
f
c
o
lo
r
-
f
il
ter
-
a
rr
a
y
ima
g
e
s
u
si
n
g
ma
c
ro
p
ixe
l
sp
e
c
tra
l
-
s
p
a
ti
a
l
tr
a
n
sf
o
rm
a
ti
o
n
”
,
I
n
P
r
o
c
.
DCC
,
p
p
.
3
-
12
,
2
0
1
2
.
[4
]
K.
H.
Ch
u
n
g
a
n
d
Y.
H.
Ch
a
n
,
“
A
fa
st
re
v
e
rs
ib
le
c
o
mp
re
ss
io
n
a
l
g
o
rit
h
m
fo
r
Ba
y
e
r
c
o
l
o
r
fi
lt
e
r
a
rr
a
y
ima
g
e
s
”
,
I
n
P
r
o
c
.
A
P
S
I
P
A
,
2
0
0
9
,
p
p
.
8
2
5
-
8
8
8
.
[5
]
C.
Do
u
tre,
e
t
a
l
.
,
“
H.
2
6
4
-
b
a
se
d
c
o
m
p
re
ss
io
n
o
f
Ba
y
e
r
p
a
tt
e
rn
v
id
e
o
se
q
u
e
n
c
e
s”
,
IEE
E
T
ra
n
s.
Circ
u
it
s
S
y
st.V
id
e
o
T
e
c
h
n
o
l
.
,
V
o
l.
1
8
,
No
.
6
,
p
p
.
7
2
5
-
7
3
4
,
Ju
n
2
0
0
8
.
[6
]
K.
H.
Ch
u
n
g
a
n
d
Y.
H.
C
h
a
n
,
“
A
lo
ss
les
s
c
o
m
p
re
ss
io
n
sc
h
e
m
e
fo
r
Ba
y
e
r
c
o
lo
r
f
il
ter
a
rra
y
i
m
a
g
e
s
”
,
IEE
E
T
ra
n
s.
Ima
g
e
Pro
c
e
ss
.
,
V
o
l.
1
7
,
N
o
.
2
,
p
p
.
1
3
4
-
1
4
4
,
F
e
b
2
0
0
8
.
[7
]
X
.
L
ian
,
e
t
a
l
.
,
“
Re
v
e
rsin
g
d
e
m
o
sa
ick
in
g
a
n
d
c
o
m
p
re
ss
io
n
i
n
c
o
lo
r
f
il
ter
a
rra
y
i
m
a
g
e
p
ro
c
e
ss
in
g
:
P
e
rf
o
rm
a
n
c
e
a
n
a
ly
sis a
n
d
m
o
d
e
li
n
g
”
,
IEE
E
T
ra
n
s.
Im
a
g
e
Pr
o
c
e
ss
.
,
V
o
l.
1
5
,
No
.
1
1
,
p
p
.
3
2
6
1
-
3
2
7
8
,
No
v
2
0
0
6
.
[8
]
R.
L
u
k
a
c
a
n
d
K.
N.
P
lata
n
io
ti
s,
“
S
in
g
le
-
se
n
so
r
c
a
m
e
ra
i
m
a
g
e
c
o
m
p
re
ss
io
n
”
,
IEE
E
T
ra
n
s.
C
o
n
su
m.
El
e
c
tro
n
,
V
o
l
.
5
2
,
N
o
.
2
,
p
p
.
2
9
9
-
3
0
7
,
M
a
y
2
0
0
6
.
[9
]
N.
Zh
a
n
g
a
n
d
X
.
L
.
W
u
,
“
L
o
ss
le
ss
c
o
m
p
re
ss
io
n
o
f
c
o
lo
r
m
o
sa
ic
i
m
a
g
e
s
”
,
IEE
E
T
ra
n
s.
Ima
g
e
Pro
c
e
ss
.
,
V
o
l.
1
5
,
N
o
.
6
,
p
p
.
1
3
7
9
-
1
3
8
8
,
Ju
n
2
0
0
6
.
[1
0
]
N.
X
.
L
ian
,
e
t
a
l.
,
“
Re
v
e
rsin
g
d
e
m
o
sa
ic
k
in
g
a
n
d
c
o
m
p
re
ss
io
n
in
c
o
lo
r
f
il
ter
a
rra
y
i
m
a
g
e
p
ro
c
e
ss
in
g
:
P
e
rf
o
rm
a
n
c
e
a
n
a
ly
sis a
n
d
m
o
d
e
li
n
g
”
,
IEE
E
T
ra
n
s.
Im
a
g
e
Pr
o
c
e
ss
.
,
V
o
l.
1
5
,
N
o
.
1
1
,
p
p
.
3
2
6
1
-
3
2
7
8
,
No
v
2
0
0
6
.
[1
1
]
B.
K.
G
u
n
tu
rk
,
e
t
a
l.
,
“
De
m
o
sa
ic
k
in
g
:
Co
lo
r
f
il
ter
a
rra
y
in
terp
o
lati
o
n
”
,
IE
EE
S
ig
n
a
l
Pro
c
e
ss
.
M
a
g
.
,
V
o
l
.
2
2
,
N
o
.
1
,
p
p
.
4
4
–
5
4
,
Ja
n
2
0
0
5
.
[1
2
]
C.
C.
Ko
h
,
e
t
a
l
.
,
“
Ne
w
e
ff
icie
n
t
m
e
th
o
d
s
o
f
i
m
a
g
e
c
o
m
p
re
ss
io
n
in
d
ig
it
a
l
c
a
m
e
ra
s
w
it
h
c
o
lo
r
f
il
ter
a
rra
y
”
,
IEE
E
T
ra
n
s.
C
o
n
s
u
m.
El
e
c
tro
n
,
Vo
l.
4
9
,
No
.
4
,
p
p
.
1
4
4
8
-
1
4
5
6
,
N
o
v
2
0
0
3
.
[1
3
]
B.
K.
G
u
n
tu
rk
,
e
t
a
l.
,
“
Co
l
o
r
p
lan
e
in
ter
p
o
lati
o
n
u
sin
g
a
lt
e
rn
a
ti
n
g
p
ro
jec
ti
o
n
s
”
,
IE
EE
T
r
a
n
s.
Im
a
g
e
Pro
c
e
ss
.
,
V
o
l
.
1
1
,
N
o
.
9
,
p
p
.
9
9
7
-
1
0
1
3
,
S
e
p
2
0
0
2
.
[1
4
]
S
.
Y.
L
e
e
a
n
d
A
.
Orte
g
a
,
“
A
n
o
v
e
l
a
p
p
r
o
a
c
h
o
f
ima
g
e
c
o
mp
re
ss
io
n
in
d
i
g
it
a
l
c
a
me
ra
s
wit
h
a
Ba
y
e
r
c
o
lo
r
fi
lt
e
r
a
rr
a
y
”
,
I
n
P
ro
c
.
I
EE
E
I
n
t.
C
o
n
f
.
I
m
a
g
e
P
ro
c
e
ss
.
,
p
p
.
4
8
2
–
4
8
5
,
Oc
t.
2
0
0
1
.
[1
5
]
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
–
JP
EG
2
0
0
0
Im
a
g
e
Co
d
in
g
S
y
ste
m
–
P
a
rt
1
:
Co
re
Co
d
i
n
g
S
y
ste
m
,
IN
CIT
S
/IS
O/IEC
S
tan
d
a
rd
1
5
4
4
4
-
1,
2
0
0
0
.
[1
6
]
X
.
W
u
a
n
d
N.
M
e
m
o
n
,
“
Co
n
tex
t
-
b
a
se
d
,
a
d
a
p
ti
v
e
,
lo
ss
les
s
ima
g
e
c
o
d
in
g
”
,
IEE
E
T
r
a
n
s.
Co
mm
u
n
.
,
V
o
l
.
4
5
,
No
.
4
,
p
p
.
4
3
7
–
4
4
4
,
A
p
r
1
9
9
7
.
[1
7
]
B.
E.
Ba
y
e
r,
“
Co
lo
r
im
a
g
in
g
a
rra
y
”
,
U.S
.
Pa
ten
t
3
9
7
1
0
6
5
,
Ju
l
1
9
7
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.