I
nte
rna
t
io
na
l J
o
urna
l o
f
Rec
o
nfig
ura
ble a
nd
E
m
be
dd
e
d Sy
s
t
e
m
s
(
I
J
R
E
S)
Vo
l.
13
,
No
.
3
,
N
o
v
e
m
b
er
20
24
,
p
p
.
5
5
2
~
5
5
9
I
SS
N:
2089
-
4864
,
DOI
:
1
0
.
1
1
5
9
1
/i
j
r
es
.
v
1
3
.
i
3
.
p
p
5
5
2
-
559
552
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
es.ia
esco
r
e.
co
m
I
m
pro
v
ed au
to
m
a
ted
pa
ra
llel i
m
pl
e
m
e
ntatio
n of
GMM
ba
ck
g
ro
und
subt
ra
ction o
n a
m
ul
ti
co
re digita
l sig
na
l proces
so
r
S
m
a
il B
a
rik
o
1
,
Ab
des
s
a
m
a
d K
lilo
u
2
,
Abdelo
ua
hed Abo
u
n
a
da
1
,
Ass
ia
Arsa
la
ne
3
1
A
u
t
o
mat
i
o
n
a
n
d
E
n
e
r
g
y
C
o
n
v
e
r
si
o
n
T
e
a
m
, F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
s a
n
d
T
e
c
h
n
o
l
o
g
i
e
s,
S
u
l
t
a
n
M
o
u
l
a
y
S
l
i
man
e
U
n
i
v
e
r
si
t
y
,
B
e
n
i
M
e
l
l
a
l
,
M
o
r
o
c
c
o
2
M
i
c
r
o
e
l
e
c
t
r
o
n
i
c
,
Em
b
e
d
d
e
d
S
y
st
e
ms a
n
d
T
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
s T
e
a
m,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
s a
n
d
T
e
c
h
n
o
l
o
g
i
e
s,
S
u
l
t
a
n
M
o
u
l
a
y
S
l
i
man
e
U
n
i
v
e
r
si
t
y
,
B
e
n
i
M
e
l
l
a
l
,
M
o
r
o
c
c
o
3
L
a
b
o
r
a
t
o
r
y
o
f
En
g
i
n
e
e
r
i
n
g
a
n
d
A
p
p
l
i
e
d
T
e
c
h
n
o
l
o
g
i
e
s,
H
i
g
h
S
c
h
o
o
l
o
f
Te
c
h
n
o
l
o
g
i
e
s,
S
u
l
t
a
n
M
o
u
l
a
y
S
l
i
man
e
U
n
i
v
e
r
si
t
y
,
B
e
n
i
M
e
l
l
a
l
,
M
o
r
o
c
c
o
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
3
0
,
2
0
2
3
R
ev
i
s
ed
Ma
r
2
8
,
2
0
2
4
A
cc
ep
ted
Ma
y
4
,
2
0
2
4
S
c
e
n
e
se
g
m
e
n
tatio
n
is
a
n
e
ss
e
n
ti
a
l
ste
p
i
n
a
w
id
e
ra
n
g
e
o
f
v
id
e
o
p
ro
c
e
ss
in
g
a
p
p
li
c
a
ti
o
n
s,
f
o
r
i
n
sta
n
c
e
,
o
b
jec
t
re
c
o
g
n
it
io
n
a
n
d
trac
k
in
g
.
T
h
e
G
a
u
ss
ian
m
i
x
tu
re
m
o
d
e
l
(G
M
M
)
f
o
r
b
a
c
k
g
ro
u
n
d
s
u
b
trac
ti
o
n
(BS
)
h
a
s
g
a
in
e
d
w
id
e
sp
re
a
d
u
sa
g
e
in
sc
e
n
e
se
g
m
e
n
tatio
n
,
d
e
s
p
it
e
it
s
k
n
o
w
n
c
o
m
p
u
tati
o
n
a
l
in
ten
sity
.
T
o
tac
k
le
th
is
c
h
a
ll
e
n
g
e
,
w
e
p
ro
p
o
se
a
p
ra
c
ti
c
a
l
s
o
lu
ti
o
n
t
o
a
c
c
e
ler
a
te
p
ro
c
e
ss
in
g
th
ro
u
g
h
a
p
a
ra
ll
e
l
im
p
le
m
e
n
tatio
n
o
n
a
n
e
m
b
e
d
d
e
d
m
u
lt
ico
re
p
latf
o
rm
.
In
th
is
p
a
p
e
r,
w
e
p
re
se
n
t
a
n
im
p
ro
v
e
d
a
u
to
m
a
ted
p
a
ra
ll
e
l
im
p
le
m
e
n
tatio
n
o
f
th
e
G
M
M
a
lg
o
rit
h
m
u
sin
g
t
h
e
Orp
h
a
n
d
irec
ti
v
e
p
ro
v
id
e
d
b
y
o
p
e
n
m
u
lt
ip
r
o
c
e
ss
in
g
(Op
e
n
M
P
).
Ex
p
e
rim
e
n
tal
a
ss
e
ss
m
e
n
ts
c
o
n
d
u
c
ted
o
n
t
h
e
e
ig
h
t
c
o
re
s
o
f
th
e
C6
6
7
8
d
ig
it
a
l
sig
n
a
l
p
ro
c
e
ss
o
r
(DS
P
)
d
e
m
o
n
stra
te
sig
n
if
ica
n
t
a
d
v
a
n
c
e
m
e
n
ts
in
p
a
ra
ll
e
l
e
ff
icie
n
c
y
,
p
a
rti
c
u
larl
y
w
h
e
n
h
a
n
d
li
n
g
h
ig
h
-
re
so
lu
ti
o
n
f
ra
m
e
s,
in
c
lu
d
i
n
g
h
ig
h
-
d
e
f
in
it
io
n
(HD
)
a
n
d
f
u
ll
-
HD
re
so
lu
ti
o
n
s.
T
h
e
a
c
h
iev
e
d
p
a
ra
ll
e
l
e
ff
icie
n
c
y
su
rp
a
ss
e
s
th
e
re
su
lt
s
o
b
tai
n
e
d
w
it
h
c
las
sic
a
l
Op
e
n
M
P
sc
h
e
d
u
li
n
g
m
o
d
e
s,
e
n
c
o
m
p
a
ss
in
g
d
y
n
a
m
ic,
sta
ti
c
,
a
n
d
g
u
id
e
d
a
p
p
ro
a
c
h
e
s.
S
p
e
c
i
f
ica
ll
y
,
th
e
p
a
ra
ll
e
l
e
f
f
icie
n
c
y
re
a
c
h
e
s
a
p
p
ro
x
im
a
tel
y
8
2
%
f
o
r
f
u
ll
-
HD
re
so
lu
ti
o
n
f
ra
m
e
s
a
n
d
,
9
9
.
3
%
f
o
r
lo
w
-
re
so
lu
ti
o
n
f
ra
m
e
s,
re
sp
e
c
ti
v
e
l
y
.
K
ey
w
o
r
d
s
:
B
ac
k
g
r
o
u
n
d
s
u
b
tr
ac
tio
n
C
6
6
7
8
DSP
Gau
s
s
ia
n
m
i
x
t
u
r
e
m
o
d
el
Op
en
m
u
ltip
r
o
ce
s
s
i
n
g
P
ar
allel
p
r
o
ce
s
s
in
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
S
m
ail
B
ar
ik
o
Au
to
m
a
tio
n
a
n
d
E
n
er
g
y
C
o
n
v
er
s
io
n
T
ea
m
,
Fac
u
lt
y
o
f
Scie
n
ce
s
an
d
T
ec
h
n
o
lo
g
ies
Su
lta
n
Mo
u
la
y
Sl
i
m
a
n
e
U
n
i
v
e
r
s
it
y
B
en
i M
ellal,
Mo
r
o
cc
o
E
m
ail:
b
ar
ik
o
_
s
m
ail
@
y
ah
o
o
.
f
r
1.
I
NT
RO
D
UCT
I
O
N
A
t
th
e
co
r
e
o
f
t
h
e
m
o
v
i
n
g
o
b
j
ec
t
d
etec
tio
n
p
r
o
ce
s
s
,
b
ac
k
g
r
o
u
n
d
s
u
b
tr
ac
tio
n
(
B
S)
is
co
n
s
id
er
ed
as
cr
itical
s
tep
r
esp
o
n
s
ib
le
f
o
r
ex
tr
ac
tin
g
m
o
v
i
n
g
o
b
j
ec
ts
[
1
]
.
Ov
er
th
e
p
ast
t
w
o
d
ec
ad
es,
th
is
f
ield
h
as
ex
p
er
ien
ce
d
s
i
g
n
i
f
ican
t
al
g
o
r
ith
m
ic
ad
v
a
n
ce
m
en
ts
,
w
i
th
r
es
ea
r
ch
er
s
p
r
o
p
o
s
in
g
v
ar
io
u
s
te
ch
n
iq
u
es
th
at
d
if
f
er
in
th
eir
ap
p
r
o
ac
h
to
u
p
d
atin
g
an
d
g
en
er
atin
g
th
e
b
ac
k
g
r
o
u
n
d
m
o
d
el.
C
o
m
p
r
eh
e
n
s
iv
e
lit
er
atu
r
e
r
ev
ie
w
s
o
f
th
ese
tec
h
n
iq
u
es
ar
e
p
r
esen
ted
b
y
Gar
cia
-
Gar
cia
et
a
l.
[
2
]
.
A
cc
o
r
d
in
g
to
[
2
]
,
BS
alg
o
r
ith
m
s
ca
n
b
e
class
i
f
ied
in
to
f
o
u
r
d
is
t
in
ct
ca
te
g
o
r
ies,
b
ased
o
n
th
e
ch
ar
ac
ter
is
t
ics
o
f
t
h
e
b
ac
k
g
r
o
u
n
d
p
atter
n
:
i)
m
at
h
e
m
a
tical
co
n
ce
p
ts
:
f
u
zz
y
m
o
d
els
[
3
]
,
s
tat
is
tical
m
o
d
el
s
[
4
]
an
d
De
m
p
s
ter
–
Sc
h
af
er
m
o
d
els
[
5
]
;
ii)
m
ac
h
i
n
e
lear
n
i
n
g
tech
n
iq
u
es:
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
s
[
6
]
a
n
d
n
eu
r
al
n
et
w
o
r
k
s
[
7
]
;
iii)
s
ig
n
al
p
r
o
ce
s
s
i
n
g
tech
n
iq
u
es:
W
ien
er
f
ilter
[
8
]
an
d
Kal
m
a
n
f
ilter
[
9
]
;
an
d
iv
)
clas
s
if
ica
tio
n
s
m
o
d
els:
cl
u
s
ter
in
g
alg
o
r
ith
m
s
[
1
0
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
I
mp
r
o
ve
d
a
u
t
o
ma
ted
p
a
r
a
llel imp
leme
n
ta
tio
n
o
f
GMM b
a
ck
g
r
o
u
n
d
s
u
b
tr
a
ctio
n
…
(
S
ma
il B
a
r
iko
)
553
W
h
ile
r
ec
en
t
BS
m
et
h
o
d
s
p
r
o
v
id
e
s
atis
f
ac
to
r
y
ac
c
u
r
ac
y
,
th
e
u
s
e
o
f
co
m
p
le
x
m
o
d
el
s
is
co
m
p
u
tatio
n
all
y
d
e
m
a
n
d
in
g
.
An
e
f
f
icien
t
al
g
o
r
ith
m
s
h
o
u
ld
b
alan
ce
ac
cu
r
ac
y
w
it
h
a
lo
w
co
m
p
u
tatio
n
al
lo
ad
.
A
cc
o
r
d
in
g
to
[
1
1
]
,
Gau
s
s
ian
m
i
x
tu
r
e
m
o
d
el
(
GM
M
)
is
a
p
r
o
m
is
i
n
g
ca
n
d
id
ate
f
o
r
ad
d
r
ess
i
n
g
t
h
ese
co
n
s
tr
ai
n
ts
,
as
it
r
ep
r
es
e
n
ts
an
o
p
ti
m
al
co
m
p
r
o
m
is
e
b
et
w
ee
n
ac
cu
r
ac
y
a
n
d
p
er
f
o
r
m
a
n
ce
.
C
o
n
s
eq
u
en
t
l
y
,
r
esear
ch
er
s
h
av
e
p
r
o
p
o
s
ed
s
tatis
tical
m
eth
o
d
s
t
o
m
a
n
a
g
e
an
d
o
v
er
co
m
e
ch
al
l
en
g
e
s
in
b
ac
k
g
r
o
u
n
d
s
ce
n
e
s
,
s
u
ch
as
i
llu
m
i
n
atio
n
an
d
n
o
is
e.
Fo
r
r
ea
l
-
w
o
r
ld
ap
p
licatio
n
s
,
r
esear
ch
er
s
h
av
e
e
m
p
lo
y
ed
b
asic
tech
n
iq
u
e
s
lik
e
GM
M
[
4
]
,
c
o
d
eb
o
o
k
[
1
0
]
an
d
v
is
u
al
b
ac
k
g
r
o
u
n
d
ex
tr
ac
to
r
(
ViB
e)
[
1
2
]
.
T
h
is
ch
o
ice
is
d
r
iv
e
n
m
a
in
l
y
b
y
m
e
m
o
r
y
an
d
ti
m
e
r
eq
u
ir
e
m
en
ts
r
eq
u
ir
ed
b
y
n
e
w
BS
m
eth
o
d
s
[
2
]
.
I
n
th
e
liter
atu
r
e,
GM
M
is
ac
k
n
o
w
led
g
ed
as
o
n
e
o
f
th
e
p
r
ed
o
m
i
n
a
n
t
s
tatis
tical
m
et
h
o
d
s
[
4
]
,
[
1
3
]
,
[
1
4
]
.
Stau
f
f
er
an
d
Gr
i
m
s
o
n
[
4
]
p
r
esen
ted
an
ad
ap
tiv
e
m
o
d
el
f
o
r
r
ea
l
-
ti
m
e
tr
ac
k
in
g
,
wh
er
e
ea
ch
p
ix
el
is
ch
ar
ac
ter
ized
b
y
a
m
ix
t
u
r
e
o
f
Gau
s
s
ia
n
s
.
T
h
is
r
ep
r
esen
tatio
n
is
d
y
n
a
m
icall
y
u
p
d
ated
in
r
ea
l
-
ti
m
e
t
h
r
o
u
g
h
t
h
e
in
co
r
p
o
r
atio
n
o
f
n
e
w
i
n
p
u
t
f
r
a
m
es.
Z
i
v
k
o
v
ic
[
1
5
]
en
h
an
c
ed
th
e
GM
M
al
g
o
r
ith
m
b
y
p
r
o
p
o
s
in
g
d
y
n
a
m
i
c
u
p
d
ates
o
f
K
Ga
u
s
s
ia
n
s
f
o
r
ea
ch
p
ix
el.
As
a
r
es
u
lt,
K
i
s
ad
ju
s
ted
d
y
n
a
m
icall
y
to
t
h
e
m
u
lt
i
m
o
d
alit
y
o
f
e
v
er
y
p
ix
el
in
ac
co
r
d
an
ce
w
it
h
s
ce
n
e
ev
o
lu
tio
n
.
T
h
e
f
ield
s
o
f
i
m
ag
e
a
n
d
v
id
e
o
p
r
o
ce
s
s
in
g
h
av
e
e
x
p
er
ien
ce
d
a
s
ig
n
i
f
ica
n
t
s
u
r
g
e
in
c
h
all
e
n
g
e
s
an
d
co
m
p
le
x
it
y
to
m
ee
t
th
e
d
e
m
an
d
s
o
f
r
ea
l
-
ti
m
e
ap
p
licatio
n
s
.
T
h
is
is
ex
p
lain
ed
b
y
th
e
m
ar
k
et
d
e
m
a
n
d
f
o
r
i
m
a
g
es
w
i
th
h
i
g
h
r
eso
l
u
tio
n
(
i.e
.
:
f
u
ll
-
h
i
g
h
-
d
ef
i
n
itio
n
(
f
u
ll
-
HD
)
1
9
2
0
×1
0
8
0
an
d
HD
1
2
8
0
×7
2
0
)
,
in
v
ar
io
u
s
ap
p
licatio
n
ar
ea
s
,
in
cl
u
d
in
g
th
e
d
etec
tio
n
o
f
tr
af
f
ic
v
i
o
latio
n
s
,
s
u
r
v
ei
llan
ce
o
f
n
at
io
n
al
b
o
r
d
er
s
,
an
d
m
o
n
ito
r
i
n
g
cr
itical
g
o
v
er
n
m
e
n
t
i
n
f
r
astru
ct
u
r
e.
C
o
n
s
eq
u
en
tl
y
,
v
id
eo
p
r
o
ce
s
s
i
n
g
h
a
s
b
ec
o
m
e
b
o
th
b
a
n
d
w
id
t
h
an
d
co
m
p
u
tatio
n
a
ll
y
i
n
te
n
s
i
v
e.
T
o
ad
d
r
ess
th
is
ch
alle
n
g
e
an
d
li
m
itat
io
n
,
p
ar
allel
p
r
o
ce
s
s
i
n
g
tec
h
n
iq
u
es
ar
e
ess
e
n
tial
to
ac
h
ie
v
e
h
i
g
h
co
m
p
u
tat
io
n
al
p
er
f
o
r
m
an
ce
an
d
f
u
l
f
ill
r
ea
l
-
ti
m
e
r
eq
u
ir
e
m
e
n
t
s
.
I
n
t
h
i
s
p
ap
er
,
th
e
co
m
p
u
tatio
n
al
p
latf
o
r
m
ch
o
s
e
n
is
th
e
m
u
ltico
r
e
C
6
6
7
8
d
ig
ital
s
ig
n
a
l
p
r
o
ce
s
s
o
r
(
DS
P
)
f
r
o
m
T
ex
as
in
s
tr
u
m
e
n
t
s
(
T
I
)
,
s
elec
ted
f
o
r
its
ad
v
an
tag
eo
u
s
f
ea
t
u
r
es,
in
cl
u
d
in
g
h
ig
h
co
m
p
u
ti
n
g
p
er
f
o
r
m
a
n
c
e
an
d
lo
w
p
o
w
er
co
n
s
u
m
p
tio
n
[
1
6
]
.
Ov
er
th
e
y
ea
r
s
,
s
ev
er
al
s
tu
d
ie
s
h
a
v
e
e
x
a
m
i
n
ed
a
u
to
m
ated
p
ar
allel
i
m
p
le
m
e
n
tatio
n
s
b
ased
o
n
o
p
en
m
u
ltip
r
o
ce
s
s
i
n
g
(
Op
en
MP
)
f
o
r
th
e
GM
M
B
S
alg
o
r
ith
m
,
w
it
h
th
e
ai
m
o
f
en
h
a
n
ci
n
g
its
co
m
p
u
ta
tio
n
al
p
er
f
o
r
m
a
n
ce
a
n
d
p
ar
allel
e
f
f
ic
ien
c
y
.
Sz
w
o
ch
et
a
l
.
[
1
7
]
s
u
g
g
ested
a
p
ar
allel
i
m
p
le
m
e
n
tatio
n
o
f
th
e
GM
M
BS
u
s
i
n
g
a
s
u
p
er
co
m
p
u
ter
co
m
p
r
is
in
g
1
9
2
n
o
d
es
co
n
n
ec
ted
w
i
th
an
I
n
f
in
iB
a
n
d
n
et
w
o
r
k
.
E
a
ch
co
m
p
u
tin
g
n
o
d
e
co
n
s
is
ted
o
f
t
w
o
s
i
x
-
co
r
e
C
P
Us
in
th
e
Xeo
n
E
M6
4
T
ar
ch
i
tectu
r
e.
Op
en
MP
w
as
u
s
ed
,
an
d
b
o
th
s
tatic
an
d
d
y
n
a
m
ic
s
ch
ed
u
li
n
g
tech
n
iq
u
es
w
er
e
e
v
alu
ated
.
T
h
e
ac
h
i
ev
ed
s
p
ee
d
u
p
v
alu
e
co
u
ld
n
o
t
ex
ce
ed
3
.
7
5
f
o
r
m
ed
iu
m
f
r
a
m
e
r
eso
l
u
tio
n
a
n
d
2
.
7
f
o
r
HD
r
eso
l
u
tio
n
f
r
a
m
es
w
h
e
n
t
w
el
v
e
th
r
e
ad
s
w
er
e
u
ti
lized
.
Ma
b
r
o
u
k
et
a
l
.
[
1
8
]
p
r
o
p
o
s
e
d
a
p
ar
allel
im
p
le
m
e
n
tatio
n
o
f
th
e
GM
M
BS
o
n
a
m
u
ltic
o
r
e
p
latf
o
r
m
,
w
h
ic
h
in
cl
u
d
ed
t
w
o
I
n
tel
Xeo
n
(
R
)
C
P
U
E
5
-
2
6
7
0
8
-
co
r
e
p
r
o
ce
s
s
o
r
s
.
T
h
e
d
is
tr
ib
u
tio
n
o
f
p
r
o
ce
s
s
i
n
g
ac
r
o
s
s
t
h
e
m
u
ltico
r
e
p
latf
o
r
m
w
a
s
ac
co
m
p
lis
h
ed
th
r
o
u
g
h
th
e
ap
p
licat
io
n
o
f
Op
en
MP
,
r
esu
ltin
g
in
a
s
p
ee
d
u
p
o
f
1
1
.
6
f
o
r
th
e
HD
r
eso
lu
tio
n
f
r
a
m
e
w
h
e
n
s
ix
tee
n
co
r
es
w
er
e
en
ab
led
.
I
n
o
u
r
p
r
ev
io
u
s
w
o
r
k
[
1
9
]
,
w
e
ev
alu
a
ted
Op
en
MP
class
ical
s
c
h
ed
u
li
n
g
(
OC
S)
m
o
d
e
s
(
e.
g
.
,
d
y
n
a
m
ic,
s
ta
tic
,
an
d
g
u
id
ed
)
,
an
d
f
o
u
n
d
t
h
at
o
n
l
y
d
y
n
a
m
ic
s
ch
ed
u
lin
g
p
r
o
v
id
ed
a
h
ig
h
s
p
ee
d
u
p
co
m
p
ar
ed
to
o
th
er
s
c
h
ed
u
li
n
g
m
o
d
es,
s
u
c
h
as
g
u
i
d
e
d
an
d
s
tatic.
T
h
e
m
ax
i
m
u
m
s
p
ee
d
u
p
ac
h
iev
ed
w
it
h
ei
g
h
t e
n
ab
led
co
r
es
w
a
s
3
.
6
f
o
r
HD
r
eso
lu
tio
n
f
r
a
m
es.
T
h
e
m
ain
co
n
tr
ib
u
tio
n
o
f
t
h
is
p
ap
er
is
th
e
p
ar
allel
e
f
f
icie
n
c
y
i
m
p
r
o
v
e
m
e
n
t
o
f
GM
M
B
S
alg
o
r
ith
m
o
n
m
u
ltico
r
e
DSP
p
latf
o
r
m
.
T
h
is
is
ac
h
ie
v
ed
b
y
s
elec
tin
g
a
s
u
itab
le
Op
en
MP
d
ir
ec
tiv
e:
Op
en
MP
o
r
p
h
an
d
ir
ec
tiv
e
(
OOD)
.
I
n
d
ee
d
,
t
h
e
OOD
ap
p
r
o
ac
h
p
r
o
v
es
p
ar
ticu
lar
l
y
ad
v
a
n
ta
g
eo
u
s
,
s
i
m
p
l
if
y
in
g
th
e
tas
k
o
f
i
m
p
le
m
en
t
in
g
co
ar
s
e
-
g
r
ai
n
p
a
r
allel
alg
o
r
it
h
m
s
[
2
0
]
,
in
w
h
i
ch
v
er
y
lar
g
e
p
r
o
g
r
a
m
r
eg
io
n
s
ar
e
p
ar
allelize
d
.
T
h
e
o
v
er
all
r
esu
lts
d
e
m
o
n
s
tr
a
te
a
s
ig
n
if
ica
n
t
i
m
p
r
o
v
e
m
en
t
in
s
p
ee
d
u
p
,
ev
en
i
n
th
e
ca
s
e
o
f
f
u
ll
-
HD
an
d
HD
r
eso
lu
tio
n
f
r
a
m
e
s
.
T
h
e
p
ap
er
i
s
s
tr
u
ct
u
r
ed
as
f
o
llo
w
s
:
se
ctio
n
2
in
tr
o
d
u
ce
s
t
h
e
GM
M
B
S
a
lg
o
r
ith
m
,
d
escr
ib
es
th
e
ex
p
er
i
m
e
n
tal
s
et
u
p
,
an
d
o
u
tli
n
es
t
h
e
p
r
o
p
o
s
ed
p
ar
allel
i
m
p
le
m
e
n
tat
io
n
ap
p
r
o
ac
h
.
Sectio
n
3
p
r
esen
ts
th
e
ex
p
er
i
m
e
n
tal
f
in
d
i
n
g
s
.
Fi
n
all
y
,
a
co
n
clu
s
io
n
is
p
r
o
v
id
ed
in
s
ec
tio
n
4
.
2.
M
AT
E
RIAL
A
ND
M
E
T
H
O
D
2
.
1
.
G
a
us
s
ia
n
m
i
x
t
ure
m
o
de
l
f
o
r
ba
ck
g
ro
un
d s
ub
t
ra
ct
io
n
GM
M
h
as
g
ai
n
ed
p
r
o
m
i
n
en
ce
in
th
e
f
ield
o
f
BS
.
T
h
e
p
io
n
ee
r
in
g
w
o
r
k
o
f
Frie
d
m
a
n
an
d
R
u
s
s
ell
[
2
1
]
in
tr
o
d
u
ce
d
a
p
r
o
b
ab
ilis
tic
m
o
d
el,
w
h
er
ei
n
ea
ch
p
i
x
el
w
a
s
ch
ar
ac
ter
ized
b
y
a
w
ei
g
h
te
d
s
u
m
o
f
a
li
m
ited
n
u
m
b
er
o
f
Gau
s
s
ian
d
is
tr
ib
u
ti
o
n
s
.
S
u
b
s
eq
u
e
n
tl
y
,
Stau
f
f
er
a
n
d
Gr
i
m
s
o
n
[
4
]
m
ad
e
s
ig
n
i
f
ic
an
t
co
n
tr
ib
u
tio
n
s
b
y
p
r
esen
tin
g
an
ad
v
a
n
ce
d
GM
M,
ac
co
m
m
o
d
ati
n
g
K
Gau
s
s
ian
m
o
d
el
s
p
er
p
ix
el,
ty
p
icall
y
K
tak
es
v
a
lu
e
w
it
h
i
n
th
e
r
an
g
e
o
f
3
to
5
[
4
]
.
T
h
is
ad
v
an
ce
m
en
t
m
ar
k
ed
a
s
ig
n
i
f
ic
an
t
s
tr
id
e
in
r
ef
in
in
g
th
e
GM
M
tech
n
iq
u
e
f
o
r
BS
.
T
h
e
f
o
r
m
u
latio
n
o
f
t
h
e
p
r
o
b
ab
ilit
y
a
s
s
o
ciate
d
w
it
h
t
h
e
c
u
r
r
en
t
p
ix
e
l
v
a
lu
e,
a
s
ill
u
s
tr
ated
in
(
1
)
,
u
n
d
er
s
co
r
es
th
e
in
h
er
e
n
t p
r
o
b
ab
ilis
tic
f
o
u
n
d
atio
n
o
f
th
i
s
m
e
th
o
d
o
lo
g
y
:
P
(
x
t
)
=
∑
ω
i,
t
*
η(
x
t
,
µ
i,
t
,
∑
i,
t
)
K
i=1
(
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
13
,
No
.
3
,
No
v
e
m
b
er
20
24
:
552
-
55
9
554
w
h
er
e:
∑
,
=
d
en
o
t
es
th
e
c
o
v
a
r
i
an
c
e
m
atr
ix
;
µ
,
=
th
e
m
ea
n
v
alu
e
;
,
=
r
ep
r
es
en
ts
th
e
w
eig
h
t
o
f
ℎ
Gau
s
s
ian
in
th
e
m
ix
tu
r
e
at
tim
e
t
;
(
,
µ
,
,
∑
,
)
=
s
p
e
cif
ie
s
th
e
Gau
s
s
i
an
p
r
o
b
a
b
i
lity
d
en
s
ity
f
u
n
ctio
n
.
C
o
m
p
ar
ativ
e
i
s
co
n
d
u
cted
b
et
w
ee
n
t
h
e
in
co
m
in
g
p
ix
e
l
an
d
th
e
GM
M
to
id
en
ti
f
y
th
e
p
ix
el
i
n
p
r
o
x
i
m
it
y
to
2
.
5
s
tan
d
ar
d
d
ev
iatio
n
s
.
T
w
o
d
is
ti
n
ct
s
ce
n
ar
io
s
ar
e
en
co
u
n
ter
ed
:
s
ce
n
a
r
io
1
:
a
m
atc
h
i
s
estab
lis
h
ed
,
p
r
o
m
p
tin
g
th
e
ad
j
u
s
t
m
e
n
t
o
f
b
o
th
th
e
m
ea
n
a
n
d
th
e
v
ar
ia
n
ce
f
o
r
th
e
co
r
r
esp
o
n
d
in
g
Gau
s
s
ia
n
d
is
tr
ib
u
tio
n
;
an
d
s
ce
n
ar
io
2
:
i
f
n
o
m
atch
is
id
en
t
if
ied
,
th
e
n
e
w
in
co
m
i
n
g
p
ix
el
s
u
b
s
tit
u
te
s
th
e
least
p
r
o
b
ab
le
co
m
p
o
n
e
n
t
w
i
th
i
n
t
h
e
m
ix
t
u
r
e
.
In
(
2
)
o
u
tlin
es t
h
e
p
r
o
ce
s
s
f
o
r
u
p
d
atin
g
th
e
w
ei
g
h
ts
o
f
th
e
K
d
is
tr
ib
u
tio
n
s
.
ω
k
,
t
=
(
1
-
α
)
ω
k
,
t
-
1
+α
(
M
k
,
t
)
(
2
)
W
h
er
e:
α
is
th
e
lear
n
in
g
r
ate
an
d
M
k
,
t
eq
u
als 1
f
o
r
th
e
m
atc
h
ed
m
o
d
el,
a
n
d
0
o
th
er
w
i
s
e.
T
h
e
p
ar
am
eter
s
u
p
d
ate
o
f
t
h
e
m
atc
h
ed
d
is
tr
ib
u
tio
n
i
s
d
ef
i
n
e
d
b
y
(
3
)
an
d
(
4
)
.
σ
2
t
=
(
1
-
ρ
)
σ
2
t
-
1
+ ρ
(
x
t
-
μ
t
)
T
(
x
t
-
μ
t
)
(
3
)
μ
t
=
(
1
-
ρ
)
μ
t
-
1
+ρ
x
t
(
4
)
W
h
er
e:
x
t
r
ep
r
esen
ts
n
e
w
i
n
p
u
t
f
r
a
m
e
p
ix
el
v
a
lu
e
;
−
1
an
d
−
1
r
ep
r
es
en
t
th
e
la
s
t
m
ea
n
a
n
d
v
ar
ian
ce
v
alu
e
s
o
f
th
e
m
atc
h
ed
Gau
s
s
ia
n
.
T
h
e
(
5
)
r
e
p
r
esen
ts
t
h
e
s
ec
o
n
d
lear
n
in
g
r
ate,
d
en
o
ted
as ρ
.
ρ=α
*
(
x
t
|
μ
k
,
σ
k
)
(
5
)
T
h
e
last
s
tep
en
co
m
p
as
s
es
t
h
e
esti
m
a
tio
n
o
f
t
h
e
b
ac
k
g
r
o
u
n
d
,
in
v
o
l
v
i
n
g
t
h
e
s
o
r
ti
n
g
o
f
Gau
s
s
ia
n
s
b
ased
o
n
th
e
ω
/σ
r
atio
.
T
h
e
in
itial
r
a
n
k
ed
d
is
tr
ib
u
t
io
n
s
B
,
w
it
h
a
cu
m
u
lati
v
e
w
eig
h
t
s
u
m
s
u
r
p
as
s
in
g
t
h
e
s
p
ec
if
ied
th
r
e
s
h
o
ld
(
Th
)
,
ar
e
id
en
tifie
d
as th
e
b
ac
k
g
r
o
u
n
d
,
as d
escr
ib
ed
b
y
th
e
(
6
)
.
B=
(
∑
ω
k
>Th
b
k
=1
)
(
6
)
W
h
er
e:
Th
r
ep
r
esen
ts
th
e
m
i
n
i
m
u
m
t
h
r
es
h
o
ld
o
f
th
e
b
ac
k
g
r
o
u
n
d
m
o
d
el.
2.
2
.
T
M
S3
2
0
C6
6
7
8
ev
a
lua
t
i
o
n
m
o
d
ule
o
v
er
v
iew
T
h
e
T
MS3
2
0
C
6
6
7
8
ev
alu
atio
n
m
o
d
u
le
w
as
u
s
ed
as
t
h
e
e
x
p
er
im
e
n
tal
p
lat
f
o
r
m
,
f
ea
t
u
r
in
g
a
s
i
n
g
le
C
6
6
7
8
DSP
ch
ip
an
d
5
1
2
MB
o
f
d
y
n
a
m
ic
r
an
d
o
m
-
ac
c
ess
m
e
m
o
r
y
(
DR
A
M
)
m
e
m
o
r
y
.
T
h
e
C
6
6
7
8
ch
ip
co
m
p
r
is
e
s
ei
g
h
t
DSP
co
r
es,
ea
ch
o
p
er
atin
g
at
a
clo
ck
f
r
eq
u
en
c
y
o
f
1
GHz
an
d
d
eli
v
er
in
g
a
co
m
p
u
ti
n
g
p
er
f
o
r
m
a
n
ce
o
f
s
i
x
tee
n
g
i
g
a
f
l
o
atin
g
-
p
o
in
t
o
p
er
atio
n
s
p
er
s
ec
o
n
d
(
GFL
OP
S
)
.
No
tab
ly
,
t
h
e
ar
ch
itectu
r
e
o
f
th
e
C
6
6
x
DSP
co
r
es
is
b
ased
o
n
v
er
y
lo
n
g
i
n
s
tr
u
ctio
n
w
o
r
d
(
VL
I
W
)
d
esig
n
[
1
6
]
,
[
2
2
]
.
T
h
e
m
e
m
o
r
y
s
tr
u
ctu
r
e
o
f
th
e
C
6
6
7
8
DSP
is
h
ier
ar
ch
ica
ll
y
o
r
g
a
n
ized
i
n
to
v
ar
io
u
s
le
v
els,
w
it
h
t
h
e
o
n
-
c
h
ip
m
e
m
o
r
y
(
L
1
)
r
ep
r
esen
ti
n
g
lev
el
1
,
en
s
u
r
in
g
e
x
p
ed
ited
C
P
U
ac
ce
s
s
co
m
p
ar
ed
to
th
e
ex
te
r
n
al
m
e
m
o
r
y
.
Deta
iled
v
ie
w
o
f
th
e
T
MS3
2
0
C
6
6
7
8
DS
P
c
o
m
p
o
n
en
t
s
is
s
h
o
w
n
i
n
Fi
g
u
r
e
1
.
A
co
m
p
r
eh
en
s
i
v
e
f
u
n
ctio
n
al
b
lo
ck
d
iag
r
a
m
o
f
th
e
C
6
6
7
8
b
o
ar
d
is
d
ep
icted
in
Fi
g
u
r
e
1
(
a)
,
w
h
ile
Fi
g
u
r
e
1
(
b
)
illu
s
tr
ates
th
e
C
6
6
7
8
ev
alu
atio
n
m
o
d
u
le.
T
h
ese
ca
p
ab
ilit
ies
h
a
v
e
i
n
s
p
ir
ed
n
u
m
er
o
u
s
r
esear
ch
co
m
m
u
n
iti
es
to
d
ev
elo
p
r
ea
l
-
ti
m
e
ap
p
licatio
n
s
u
s
i
n
g
th
is
h
ar
d
w
ar
e
p
latf
o
r
m
[
2
3
]
–
[
2
6
]
.
T
h
r
o
u
g
h
o
u
t
o
u
r
i
m
p
le
m
e
n
tatio
n
,
w
e
u
t
ilized
v
er
s
io
n
8
.
3
.
7
o
f
th
e
C
6
0
0
0
T
I
co
m
p
iler
.
2.
3
.
P
a
ra
lleliza
t
io
n m
et
ho
d
T
h
e
Op
en
MP
s
er
v
es
as
a
n
ap
p
licatio
n
p
r
o
g
r
am
m
i
n
g
i
n
ter
f
ac
e
(
A
P
I
)
th
at
f
ac
ilit
at
es
p
ar
allel
p
r
o
g
r
am
m
i
n
g
o
n
m
u
lt
ico
r
e
p
latf
o
r
m
s
c
h
ar
ac
ter
ized
b
y
h
o
m
o
g
en
eo
u
s
p
r
o
ce
s
s
o
r
s
an
d
s
h
ar
ed
m
e
m
o
r
y
ar
ch
itect
u
r
es
[
2
7
]
,
[
2
8
]
.
I
t
f
a
cilitates
th
e
h
a
n
d
li
n
g
o
f
p
ar
allel
i
m
p
le
m
e
n
tat
io
n
s
b
y
o
f
f
er
in
g
d
ir
ec
tiv
e
s
th
a
t
s
p
ec
if
y
to
t
h
e
co
m
p
iler
t
h
e
p
ar
allel
r
eg
io
n
s
w
it
h
i
n
t
h
e
co
d
e.
User
s
ar
e
also
r
eq
u
ir
ed
to
s
elec
t
ap
p
r
o
p
r
iate
s
ch
ed
u
lin
g
tech
n
iq
u
es
to
ef
f
ec
tiv
el
y
d
i
s
tr
ib
u
te
p
r
o
ce
s
s
in
g
task
s
a
m
o
n
g
d
if
f
er
e
n
t
co
r
es.
T
h
e
ch
o
ice
o
f
s
ch
ed
u
lin
g
t
y
p
e
s
i
g
n
if
ica
n
tl
y
in
f
l
u
e
n
ce
s
th
e
o
v
er
all
p
er
f
o
r
m
an
ce
o
u
tco
m
e
s
.
On
th
e
o
th
er
h
an
d
,
a
d
ee
p
u
n
d
er
s
ta
n
d
in
g
o
f
t
h
e
alg
o
r
ith
m
s
tr
u
c
tu
r
e
an
d
th
e
n
a
tu
r
e
o
f
th
e
alg
o
r
it
h
m
's
w
o
r
k
lo
ad
lo
o
p
is
co
n
s
id
er
ed
a
k
ey
f
ac
to
r
in
id
en
tify
i
n
g
ac
cu
r
ate
Op
en
MP
s
ch
ed
u
lin
g
.
I
n
d
ee
d
,
in
th
e
ca
s
e
o
f
th
e
GM
M
B
S
alg
o
r
ith
m
,
th
e
w
o
r
k
lo
ad
is
co
n
s
id
e
r
ed
ir
r
eg
u
lar
.
T
h
is
ir
r
eg
u
lar
it
y
ar
is
e
s
f
r
o
m
t
h
e
d
y
n
a
m
ic
n
at
u
r
e
o
f
th
e
alg
o
r
ith
m
a
n
d
its
d
ep
en
d
en
ce
o
n
th
e
co
m
p
le
x
it
y
o
f
t
h
e
s
ce
n
e.
T
h
e
ir
r
eg
u
lar
w
o
r
k
lo
ad
ca
n
b
e
attr
ib
u
ted
to
s
ev
er
al
f
ac
to
r
s
,
in
cl
u
d
in
g
:
i)
v
ar
y
i
n
g
b
ac
k
g
r
o
u
n
d
co
m
p
le
x
it
y
:
Dif
f
er
en
t
p
ix
els
i
n
an
i
m
ag
e
m
a
y
h
a
v
e
v
ar
y
in
g
co
m
p
le
x
itie
s
in
th
e
b
ac
k
g
r
o
u
n
d
d
u
e
to
c
h
an
g
es
in
li
g
h
ti
n
g
o
r
o
b
j
ec
t
m
o
v
e
m
en
ts
a
n
d
ii)
ad
ap
tiv
e
m
o
d
el
u
p
d
atin
g
:
GM
M
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
I
mp
r
o
ve
d
a
u
t
o
ma
ted
p
a
r
a
llel imp
leme
n
ta
tio
n
o
f
GMM b
a
ck
g
r
o
u
n
d
s
u
b
tr
a
ctio
n
…
(
S
ma
il B
a
r
iko
)
555
m
o
d
el
s
n
ee
d
to
b
e
co
n
ti
n
u
o
u
s
l
y
u
p
d
ated
b
ased
o
n
th
e
ch
ar
a
cter
is
tics
o
f
t
h
e
s
ce
n
e.
P
ix
el
s
w
it
h
m
o
r
e
d
y
n
a
m
ic
b
ac
k
g
r
o
u
n
d
s
o
r
v
ar
iatio
n
s
w
o
u
ld
r
eq
u
ir
e
m
o
r
e
f
r
eq
u
e
n
t u
p
d
ates,
r
esu
lt
in
g
i
n
a
h
ea
v
ier
w
o
r
k
lo
ad
.
(
a)
(
b
)
Fig
u
r
e
1
.
Ov
er
v
ie
w
o
f
th
e
T
MS3
2
0
C
6
6
7
8
DSP
(
a)
f
u
n
ctio
n
a
l b
lo
ck
d
iag
r
a
m
o
f
t
h
e
T
MS3
2
0
C
6
6
7
8
DSP an
d
(
b
)
E
VM
6
6
7
8
d
ev
elo
p
m
e
n
t b
o
ar
d
Du
e
to
t
h
e
ir
r
eg
u
lar
n
at
u
r
e
o
f
t
h
e
GM
M
alg
o
r
it
h
m
,
w
e
c
h
o
s
e
t
h
e
OO
D
ap
p
r
o
ac
h
,
w
h
ich
o
f
f
er
s
s
ig
n
i
f
ica
n
t
ad
v
an
tag
e
s
in
s
i
m
p
lify
in
g
th
e
i
m
p
le
m
en
tatio
n
o
f
co
ar
s
e
-
g
r
ai
n
p
ar
allel
alg
o
r
ith
m
s
[
2
0
]
.
C
o
ar
s
e
-
g
r
ain
ed
p
ar
allelis
m
p
r
o
v
es
to
b
e
a
s
u
itab
le
s
tr
ate
g
y
f
o
r
ir
r
eg
u
lar
lo
o
p
alg
o
r
it
h
m
s
,
w
h
er
e
th
e
w
o
r
k
lo
ad
p
er
iter
atio
n
ex
h
ib
its
s
u
b
s
ta
n
tial
v
ar
iatio
n
s
.
T
h
is
ap
p
r
o
ac
h
en
tails
d
iv
id
in
g
t
h
e
o
v
er
all
task
in
to
lar
g
er
u
n
its
o
f
w
o
r
k
,
w
it
h
ea
ch
u
n
i
t
r
ep
r
esen
ti
n
g
a
s
i
g
n
if
ican
t
p
o
r
tio
n
o
f
th
e
to
tal
w
o
r
k
lo
ad
.
C
o
ar
s
e
-
g
r
ain
ed
p
ar
allelis
m
ef
f
ec
tiv
e
l
y
ac
co
m
m
o
d
ates
th
e
ir
r
eg
u
lar
itie
s
i
n
w
o
r
k
lo
ad
,
r
ed
u
cin
g
s
y
n
c
h
r
o
n
izat
io
n
o
v
e
r
h
ea
d
co
m
p
ar
ed
to
f
i
n
e
-
g
r
ai
n
ed
ap
p
r
o
ac
h
es.
T
h
is
m
ak
e
s
it
p
ar
ticu
lar
l
y
ef
f
ec
tiv
e
f
o
r
s
ce
n
ar
io
s
w
h
er
e
th
e
co
m
p
u
tatio
n
al
r
eq
u
ir
e
m
en
ts
o
f
d
if
f
er
en
t
iter
atio
n
s
v
ar
y
w
id
el
y
.
Ov
er
all,
th
e
OOD
e
m
p
o
w
er
s
u
s
er
s
w
ith
m
o
r
e
n
u
an
ce
d
co
n
tr
o
l
o
v
er
p
ar
alleliza
tio
n
,
l
ea
d
in
g
to
en
h
a
n
ce
d
p
er
f
o
r
m
a
n
ce
a
n
d
i
m
p
r
o
v
ed
s
tab
ilit
y
i
n
p
ar
allel
p
r
o
g
r
a
m
s
,
esp
ec
iall
y
i
n
ca
s
e
s
w
h
er
e
n
e
s
t
ed
p
ar
allel
r
eg
io
n
s
ar
e
in
v
o
l
v
e
d
.
T
h
e
A
lg
o
r
it
h
m
1
s
h
o
w
s
t
h
e
p
s
eu
d
o
co
d
e
o
f
o
u
r
p
r
o
p
o
s
ed
i
m
p
le
m
e
n
tat
io
n
u
s
i
n
g
OO
D
ap
p
r
o
ac
h
.
I
n
th
is
ca
s
e,
t
h
e
“
o
m
p
f
o
r
”
d
ir
ec
tiv
e
i
n
B
ac
k
g
r
o
u
n
d
_
Su
b
tr
a
cto
r
GM
M
f
u
n
ctio
n
is
co
n
s
id
er
ed
as
an
o
r
p
h
an
d
ir
ec
tiv
e.
T
h
e
u
tili
za
tio
n
o
f
an
o
r
p
h
an
d
ir
ec
tiv
e
in
th
is
co
n
te
x
t
h
i
g
h
lig
h
t
s
a
k
e
y
asp
ec
t
o
f
o
u
r
d
esig
n
s
tr
ateg
y
,
e
m
p
h
a
s
izi
n
g
t
h
e
p
ar
alleliza
ti
o
n
o
f
th
e
BS
p
r
o
ce
s
s
.
Fi
g
u
r
e
2
s
h
o
w
s
C
D
n
et
2
0
1
2
h
ig
h
w
a
y
s
u
b
-
d
ataset.
Fig
u
r
e
2
(
a)
p
r
esen
ts
a
v
is
u
al
r
ep
r
esen
tatio
n
o
f
t
h
e
in
p
u
t
f
r
a
m
e,
p
r
o
v
id
in
g
a
clea
r
s
n
ap
s
h
o
t
o
f
th
e
r
a
w
d
ata
p
r
o
ce
s
s
ed
f
r
o
m
C
D
n
et
d
atase
t
[
2
9
]
.
A
d
d
itio
n
all
y
,
Fi
g
u
r
e
2
(
b
)
co
m
p
le
m
e
n
ts
th
i
s
b
y
ill
u
s
tr
atin
g
t
h
e
g
en
er
ated
m
as
k
,
d
er
iv
ed
f
r
o
m
o
u
r
e
n
h
a
n
ce
d
p
ar
allel
im
p
le
m
en
tatio
n
o
f
th
e
GM
M
al
g
o
r
ith
m
.
A
l
g
o
r
ith
m
1
.
P
ar
allel
i
m
p
le
m
e
n
tatio
n
o
f
GM
M
B
S u
s
in
g
OO
D
ap
p
r
o
ac
h
//Main function
1.
omp_set_num_threads (8);
2.
GmmModel
defined by {Mean(mk), Weight(ωk) and variance(σk);
3.
Perform all GMM model initialization;
4.
Get current input frame;
5.
Begin
6.
#pragma omp parallel
7.
Call Background_SubtractorGMM (In InputFrame,
inout GMMModel,
Out MaskFrame);
8.
end function
//Background_SubtractorGMM function
1.
#pragma omp for
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
13
,
No
.
3
,
No
v
e
m
b
er
20
24
:
552
-
55
9
556
2.
For each (PixelIndex: = [0…SizeOfFrame [; PixelIndex+
+)
3.
Begin
4.
pixel= InputFrame[PixelIndex];
5.
For each k Gaussian
6.
Begin
7.
diff(k) = abs (mk
-
pixel);
8.
if (diff[k] < Tmatch) then
9.
Update GMMModel {Mean(mk), Weight(ωk) and variance(σk)};
10.
else
11.
Up
date GMMModel {Weight(ωk)};
12.
end if
13.
end for
14.
Normalization of Weight (ωk).
15.
For each k Gaussian
16.
Begin
17.
Rank and sort all Gaussians by the ratio ωk ⁄σk;
18.
end for
19.
Retain the first B componants
whose weight is greater than threshold (Th);
20.
if (pixel does not match background model) then
21.
mark pixel in MaskFrame as foreground;
22.
else
23.
mark pixel in MaskFrame as background;
24.
end for
25.
end function
(
a)
(
b
)
Fig
u
r
e
2
.
C
Dn
et
2
0
1
2
h
ig
h
w
a
y
s
u
b
-
d
ataset
o
f
(
a)
in
p
u
t
f
r
a
m
e
an
d
(
b
)
g
en
er
ate
m
as
k
[
2
9
]
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
n
e
w
p
ar
allel
ap
p
r
o
ac
h
p
r
o
p
o
s
ed
in
th
is
p
ap
er
,
b
ased
o
n
OOD,
w
as
ap
p
lied
to
d
if
f
er
en
t
f
r
a
m
e
r
eso
lu
tio
n
s
to
v
alid
ate
t
h
e
e
f
f
ec
tiv
e
n
ess
o
f
t
h
is
m
et
h
o
d
.
T
h
e
r
esu
lt
s
o
b
tain
ed
f
o
r
v
ar
io
u
s
r
eso
lu
tio
n
f
r
a
m
es
(
Fig
u
r
es
3
an
d
4
)
,
as
p
r
ese
n
te
d
in
Fi
g
u
r
e
3
(
a)
f
o
r
lo
w
-
r
eso
l
u
tio
n
f
r
a
m
es,
Fig
u
r
e
3
(
b
)
f
o
r
m
ed
iu
m
-
r
eso
l
u
tio
n
f
r
a
m
e
s
,
Fig
u
r
e
4
(
a)
f
o
r
HD
f
r
a
m
e
s
,
an
d
Fig
u
r
e
4
(
b
)
f
o
r
f
u
ll
-
HD
f
r
a
m
es,
d
e
m
o
n
s
tr
ate
th
at
o
u
r
p
r
o
p
o
s
ed
m
et
h
o
d
y
ield
s
i
m
p
r
o
v
ed
r
esu
lts
co
m
p
ar
ed
to
OC
S
m
o
d
es
p
r
esen
ted
in
p
r
io
r
w
o
r
k
[
1
9
]
.
T
h
e
d
y
n
a
m
i
c
s
ch
ed
u
lin
g
w
it
h
a
ch
u
n
k
s
ize
eq
u
al
to
1
2
8
p
r
o
v
id
es
th
e
b
est
s
p
ee
d
u
p
r
esu
lts
,
as
p
r
ese
n
ted
in
[
1
9
]
.
Ho
w
e
v
er
,
in
th
e
c
u
r
r
en
t
w
o
r
k
,
o
u
r
n
e
w
a
p
p
r
o
ac
h
b
ased
o
n
OOD
o
u
tp
er
f
o
r
m
s
t
h
e
O
C
S
m
et
h
o
d
s
.
L
i
n
ea
r
s
p
ee
d
u
p
w
a
s
ac
h
ie
v
ed
f
o
r
th
e
lo
w
-
r
eso
l
u
tio
n
f
r
a
m
e,
as
s
h
o
w
n
i
n
Fi
g
u
r
e
3
(
a)
.
Ho
w
e
v
er
,
w
e
o
b
s
er
v
ed
a
d
ec
r
ea
s
e
in
s
p
ee
d
u
p
f
o
r
m
ed
i
u
m
r
eso
lu
tio
n
f
r
a
m
e,
as
illu
s
tr
ated
in
Fig
u
r
e
3
(
b
)
,
s
tar
tin
g
f
r
o
m
th
e
s
ev
e
n
t
h
co
r
e.
Fo
r
HD
an
d
f
u
ll
-
HD
f
r
a
m
es,
a
s
p
ee
d
u
p
d
ec
r
ea
s
e
w
as
n
o
ticed
w
h
e
n
u
s
in
g
m
o
r
e
th
an
s
i
x
ac
tiv
ated
co
r
es
a
s
ill
u
s
tr
ated
Fig
u
r
e
4
(
a)
an
d
F
ig
u
r
e
4
(
b
)
,
r
esp
ec
t
iv
el
y
.
T
h
is
r
ed
u
ct
io
n
i
n
s
p
ee
d
u
p
ca
n
b
e
attr
ib
u
ted
to
th
e
l
i
m
itatio
n
o
f
th
e
D
R
A
M
m
e
m
o
r
y
b
an
d
w
id
t
h
.
A
cc
es
s
to
t
h
e
DR
AM
is
r
e
s
tr
icted
to
a
s
i
n
g
le
co
r
e
at
a
tim
e,
u
tili
zi
n
g
a
6
4
-
b
it in
ter
f
ac
e
[
1
6
]
.
B
y
s
tr
ateg
ica
ll
y
a
lig
n
i
n
g
th
e
OOD
w
it
h
th
e
s
p
ec
if
ic
d
e
m
an
d
s
o
f
t
h
e
GM
M
B
S
al
g
o
r
ith
m
,
w
e
s
u
cc
ee
d
ed
in
o
p
ti
m
iz
in
g
t
h
e
alg
o
r
ith
m
's
p
ar
alleliza
t
io
n
.
A
s
s
h
o
w
n
i
n
Fi
g
u
r
e
5
,
th
e
OO
D
p
r
o
v
id
es
th
e
b
est
p
ar
allel
ef
f
icie
n
c
y
p
er
f
o
r
m
an
ce
co
m
p
ar
ed
to
th
e
co
n
v
e
n
ti
o
n
al
Op
en
MP
s
ch
ed
u
l
in
g
m
e
th
o
d
s
p
r
esen
ted
i
n
[
1
9
]
.
T
h
e
ad
o
p
ti
o
n
o
f
t
h
e
OO
D,
s
u
b
s
eq
u
en
t
to
co
d
e
r
ea
llo
ca
tio
n
,
r
ep
r
esen
ts
a
s
tr
ate
g
ic
m
o
v
e
to
b
o
ls
ter
t
h
e
p
ar
allel
ef
f
icie
n
c
y
o
f
t
h
e
GM
M
B
S a
lg
o
r
ith
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
I
mp
r
o
ve
d
a
u
t
o
ma
ted
p
a
r
a
llel imp
leme
n
ta
tio
n
o
f
GMM b
a
ck
g
r
o
u
n
d
s
u
b
tr
a
ctio
n
…
(
S
ma
il B
a
r
iko
)
557
(
a)
(
b
)
Fig
u
r
e
3
.
C
o
m
p
ar
is
o
n
o
f
o
b
tai
n
ed
s
p
ee
d
u
p
b
et
w
ee
n
OOD
a
n
d
OC
S
ap
p
r
o
ac
h
es in
(
a)
3
2
0
×
2
4
0
f
r
am
e
r
eso
lu
tio
n
a
n
d
(
b
)
6
4
0
×4
8
0
f
r
a
m
e
r
eso
lu
t
io
n
(
a)
(
b
)
Fig
u
r
e
4
.
C
o
m
p
ar
is
o
n
o
f
o
b
tai
n
ed
s
p
ee
d
u
p
b
et
w
ee
n
OOD
a
n
d
OC
S a
p
p
r
o
ac
h
es in
(
a)
1
2
8
0
×7
2
0
f
r
am
e
r
eso
lu
tio
n
a
n
d
(
b
)
1
9
2
0
×1
0
8
0
f
r
a
m
e
r
eso
l
u
tio
n
Fig
u
r
e
5
.
P
ar
allel
ef
f
icie
n
c
y
c
o
m
p
ar
is
o
n
b
et
w
ee
n
O
OD
an
d
OC
S
ap
p
r
o
ac
h
es
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
13
,
No
.
3
,
No
v
e
m
b
er
20
24
:
552
-
55
9
558
4.
CO
NCLU
SI
O
N
I
n
th
is
p
ap
er
,
w
e
d
e
m
o
n
s
tr
a
ted
th
e
p
ar
allel
i
m
p
le
m
en
tat
io
n
o
f
th
e
GM
M
BS
al
g
o
r
i
th
m
o
n
a
m
u
ltico
r
e
DSP
p
lat
f
o
r
m
u
s
i
n
g
Op
en
MP
.
Af
ter
co
n
d
u
c
ti
n
g
a
co
m
p
r
eh
e
n
s
iv
e
an
al
y
s
is
o
f
t
h
e
GM
M
B
S
alg
o
r
ith
m
's
w
o
r
k
lo
ad
an
d
s
tr
u
ctu
r
al
i
n
tr
icac
ies,
w
e
r
ec
o
g
n
iz
ed
th
e
n
ee
d
f
o
r
an
ad
ap
tiv
e
a
p
p
r
o
ac
h
d
u
e
to
th
e
ir
r
eg
u
lar
w
o
r
k
lo
ad
p
er
p
ix
el
p
r
o
ce
s
s
in
g
.
I
n
t
h
is
co
n
te
x
t,
t
h
e
in
te
g
r
atio
n
o
f
t
h
e
o
r
p
h
an
d
ir
ec
tiv
e
f
r
o
m
th
e
Op
en
MP
A
P
I
p
lay
ed
a
cr
u
cial
r
o
le
in
ac
h
iev
in
g
o
p
ti
m
al
p
er
f
o
r
m
a
n
ce
,
s
u
r
p
ass
i
n
g
alter
n
ativ
e
s
c
h
ed
u
l
in
g
m
o
d
e
s
s
u
c
h
as
d
y
n
a
m
ic,
s
ta
t
ic,
an
d
g
u
id
ed
.
I
n
d
ee
d
,
w
e
en
h
a
n
ce
d
t
h
e
GM
M
B
S
al
g
o
r
ith
m
's
p
r
o
ce
s
s
i
n
g
ca
p
ab
ilit
ies,
r
esu
lti
n
g
in
s
i
g
n
if
ica
n
t
i
m
p
r
o
v
e
m
en
ts
i
n
p
ar
a
llel
ef
f
ic
ien
c
y
.
Sp
ec
i
f
icall
y
,
w
e
ac
h
ie
v
ed
8
2
%
p
ar
allel
ef
f
ic
ien
c
y
f
o
r
f
u
l
l
-
HD
r
eso
lu
tio
n
f
r
a
m
e
a
n
d
a
li
n
ea
r
s
p
ee
d
u
p
(
i.e
.
,
9
9
.
3
%
p
ar
allel
e
f
f
icien
c
y
)
f
o
r
lo
w
-
r
eso
lu
tio
n
f
r
a
m
e
w
h
e
n
all
ei
g
h
t
DSP
co
r
es
w
er
e
e
n
ab
led
.
L
o
o
k
in
g
ah
ea
d
,
o
u
r
f
u
t
u
r
e
w
o
r
k
ai
m
s
to
e
x
p
an
d
th
e
p
ar
allel
i
m
p
le
m
en
ta
tio
n
o
f
th
e
GM
M
B
S
alg
o
r
it
h
m
to
1
6
DS
P
co
r
es,
f
o
llo
w
ed
b
y
t
h
e
p
ar
allel
i
m
p
le
m
e
n
tatio
n
o
f
v
e
h
icle
tr
ac
k
i
n
g
p
r
o
ce
s
s
in
g
ch
ain
.
RE
F
E
R
E
NC
E
S
[
1
]
S.
-
C
.
S
.
C
h
e
u
n
g
a
n
d
C
.
K
a
mat
h
,
“
R
o
b
u
s
t
b
a
c
k
g
r
o
u
n
d
s
u
b
t
r
a
c
t
i
o
n
w
i
t
h
f
o
r
e
g
r
o
u
n
d
v
a
l
i
d
a
t
i
o
n
f
o
r
u
r
b
a
n
t
r
a
f
f
i
c
v
i
d
e
o
,
”
EU
R
AS
I
P
J
o
u
rn
a
l
o
n
A
d
v
a
n
c
e
s
i
n
S
i
g
n
a
l
Pr
o
c
e
s
si
n
g
,
v
o
l
.
2
0
0
5
,
n
o
.
1
4
,
p
.
7
2
6
2
6
1
,
D
e
c
.
2
0
0
5
,
d
o
i
:
1
0
.
1
1
5
5
/
A
S
P
.
2
0
0
5
.
2
3
3
0
.
[
2
]
B
.
G
a
r
c
i
a
-
G
a
r
c
i
a
,
T
.
B
o
u
w
man
s
,
a
n
d
A
.
J.
R
.
S
i
l
v
a
,
“
B
a
c
k
g
r
o
u
n
d
s
u
b
t
r
a
c
t
i
o
n
i
n
r
e
a
l
a
p
p
l
i
c
a
t
i
o
n
s:
c
h
a
l
l
e
n
g
e
s,
c
u
r
r
e
n
t
mo
d
e
l
s
a
n
d
f
u
t
u
r
e
d
i
r
e
c
t
i
o
n
s
,
”
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
Re
v
i
e
w
,
v
o
l
.
3
5
,
p
.
1
0
0
2
0
4
,
F
e
b
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
sr
e
v
.
2
0
1
9
.
1
0
0
2
0
4
.
[
3
]
W
.
K
i
m
a
n
d
C
.
K
i
m,
“
B
a
c
k
g
r
o
u
n
d
s
u
b
t
r
a
c
t
i
o
n
f
o
r
d
y
n
a
mi
c
t
e
x
t
u
r
e
sce
n
e
s
u
si
n
g
f
u
z
z
y
c
o
l
o
r
h
i
st
o
g
r
a
ms,”
I
EE
E
S
i
g
n
a
l
P
ro
c
e
ss
i
n
g
L
e
t
t
e
rs
,
v
o
l
.
1
9
,
n
o
.
3
,
p
p
.
1
2
7
–
1
3
0
,
M
a
r
.
2
0
1
2
,
d
o
i
:
1
0
.
1
1
0
9
/
L
S
P
.
2
0
1
1
.
2
1
8
2
6
4
8
.
[
4
]
C
.
S
t
a
u
f
f
e
r
a
n
d
W
.
E.
L
.
G
r
i
mso
n
,
“
A
d
a
p
t
i
v
e
b
a
c
k
g
r
o
u
n
d
mi
x
t
u
r
e
mo
d
e
l
s
f
o
r
r
e
a
l
-
t
i
me
t
r
a
c
k
i
n
g
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s.
1
9
9
9
I
EE
E
C
o
m
p
u
t
e
r
S
o
c
i
e
t
y
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
s
i
o
n
a
n
d
P
a
t
t
e
r
n
Re
c
o
g
n
i
t
i
o
n
(
C
a
t
.
N
o
PR
0
0
1
4
9
)
,
p
p
.
2
4
6
–
2
5
2
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
.
1
9
9
9
.
7
8
4
6
3
7
.
[
5
]
O
.
M
u
n
t
e
a
n
u
,
T
.
B
o
u
w
man
s
,
E.
-
H
.
Z
a
h
z
a
h
,
a
n
d
R
.
V
a
s
i
u
,
“
T
h
e
d
e
t
e
c
t
i
o
n
o
f
mo
v
i
n
g
o
b
j
e
c
t
s
i
n
v
i
d
e
o
b
y
b
a
c
k
g
r
o
u
n
d
s
u
b
t
r
a
c
t
i
o
n
u
si
n
g
D
e
mp
st
e
r
-
S
h
a
f
e
r
t
h
e
o
r
y
,
”
Ed
i
t
u
ra
P
o
l
i
t
e
h
n
i
c
a
,
2
0
1
5
.
[
6
]
H
.
-
H
.
L
i
n
,
T
.
-
L
.
L
i
u
,
a
n
d
J
.
-
H
.
C
h
u
a
n
g
,
“
A
p
r
o
b
a
b
i
l
i
s
t
i
c
S
V
M
a
p
p
r
o
a
c
h
f
o
r
b
a
c
k
g
r
o
u
n
d
s
c
e
n
e
i
n
i
t
i
a
l
i
z
a
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s.
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
I
m
a
g
e
P
ro
c
e
ssi
n
g
,
v
o
l
.
1
,
p
p
.
8
9
3
–
8
9
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
P
.
2
0
0
2
.
1
0
3
9
1
1
6
.
[
7
]
D
.
C
u
l
i
b
r
k
,
O
.
M
a
r
q
u
e
s,
D
.
S
o
c
e
k
,
H
.
K
a
l
v
a
,
a
n
d
B
.
F
u
r
h
t
,
“
N
e
u
r
a
l
n
e
t
w
o
r
k
a
p
p
r
o
a
c
h
t
o
b
a
c
k
g
r
o
u
n
d
mo
d
e
l
i
n
g
f
o
r
v
i
d
e
o
o
b
j
e
c
t
se
g
me
n
t
a
t
i
o
n
,
”
I
E
EE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
N
e
u
r
a
l
N
e
t
w
o
rks
,
v
o
l
.
1
8
,
n
o
.
6
,
p
p
.
1
6
1
4
–
1
6
2
7
,
N
o
v
.
2
0
0
7
,
d
o
i
:
1
0
.
1
1
0
9
/
T
N
N
.
2
0
0
7
.
8
9
6
8
6
1
.
[
8
]
K
.
T
o
y
a
ma,
J.
K
r
u
mm
,
B
.
B
r
u
m
i
t
t
,
a
n
d
B
.
M
e
y
e
r
s,
“
W
a
l
l
f
l
o
w
e
r
:
p
r
i
n
c
i
p
l
e
s
a
n
d
p
r
a
c
t
i
c
e
o
f
b
a
c
k
g
r
o
u
n
d
mai
n
t
e
n
a
n
c
e
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
S
e
v
e
n
t
h
I
EE
E
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
si
o
n
,
1
9
9
9
,
p
p
.
2
5
5
–
2
6
1
v
o
l
.
1
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
V
.
1
9
9
9
.
7
9
1
2
2
8
.
[
9
]
S
.
M
e
sse
l
o
d
i
,
C
.
M
.
M
o
d
e
n
a
,
N
.
S
e
g
a
t
a
,
a
n
d
M
.
Z
a
n
i
n
,
“
A
K
a
l
man
f
i
l
t
e
r
b
a
se
d
b
a
c
k
g
r
o
u
n
d
u
p
d
a
t
i
n
g
a
l
g
o
r
i
t
h
m
r
o
b
u
st
t
o
sh
a
r
p
i
l
l
u
mi
n
a
t
i
o
n
c
h
a
n
g
e
s,”
I
n
:
R
o
l
i
,
F
.
,
V
i
t
u
l
a
n
o
,
S
.
(
e
d
s)
I
m
a
g
e
A
n
a
l
y
si
s
a
n
d
Pro
c
e
ssi
n
g
–
I
C
I
AP
2
0
0
5
.
I
C
I
AP
2
0
0
5
.
L
e
c
t
u
re
N
o
t
e
s
i
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
3
6
1
7
.
S
p
r
i
n
g
e
r
,
B
e
r
l
i
n
,
H
e
i
d
e
l
b
e
r
g
,
2
0
0
5
,
d
o
i
:
1
0
.
1
0
0
7
/
1
1
5
5
3
5
9
5
_
2
0
.
[
1
0
]
K
.
K
i
m,
T
.
H
.
C
h
a
l
i
d
a
b
h
o
n
g
se
,
D
.
H
a
r
w
o
o
d
,
a
n
d
L
.
D
a
v
i
s,
“
B
a
c
k
g
r
o
u
n
d
m
o
d
e
l
i
n
g
a
n
d
su
b
t
r
a
c
t
i
o
n
b
y
c
o
d
e
b
o
o
k
c
o
n
st
r
u
c
t
i
o
n
,
”
i
n
2
0
0
4
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
2
0
0
4
.
I
C
I
P
’
0
4
.
,
v
o
l
.
5
,
p
p
.
3
0
6
1
–
3
0
6
4
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
P
.
2
0
0
4
.
1
4
2
1
7
5
9
.
[
1
1
]
Y
.
B
e
n
e
z
e
t
h
,
P
.
M
.
Jo
d
o
i
n
,
B
.
Em
i
l
e
,
H
.
L
a
u
r
e
n
t
,
a
n
d
C
.
R
o
se
n
b
e
r
g
e
r
,
“
R
e
v
i
e
w
a
n
d
e
v
a
l
u
a
t
i
o
n
o
f
c
o
mm
o
n
l
y
-
i
mp
l
e
m
e
n
t
e
d
b
a
c
k
g
r
o
u
n
d
s
u
b
t
r
a
c
t
i
o
n
a
l
g
o
r
i
t
h
ms,
”
i
n
2
0
0
8
1
9
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Pa
t
t
e
rn
R
e
c
o
g
n
i
t
i
o
n
,
D
e
c
.
2
0
0
8
,
p
p
.
1
–
4
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
P
R
.
2
0
0
8
.
4
7
6
0
9
9
8.
[
1
2
]
O
.
B
a
r
n
i
c
h
a
n
d
M
.
V
.
D
r
o
o
g
e
n
b
r
o
e
c
k
,
“
V
i
B
E:
A
p
o
w
e
r
f
u
l
r
a
n
d
o
m
t
e
c
h
n
i
q
u
e
t
o
e
st
i
mat
e
t
h
e
b
a
c
k
g
r
o
u
n
d
i
n
v
i
d
e
o
se
q
u
e
n
c
e
s,”
i
n
2
0
0
9
I
EEE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Ac
o
u
st
i
c
s,
S
p
e
e
c
h
a
n
d
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
,
A
p
r
.
2
0
0
9
,
p
p
.
9
4
5
–
9
4
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
S
S
P
.
2
0
0
9
.
4
9
5
9
7
4
1
.
[
1
3
]
M
.
H
o
f
ma
n
n
,
P
.
T
i
e
f
e
n
b
a
c
h
e
r
,
a
n
d
G
.
R
i
g
o
l
l
,
“
B
a
c
k
g
r
o
u
n
d
se
g
me
n
t
a
t
i
o
n
w
i
t
h
f
e
e
d
b
a
c
k
:
t
h
e
p
i
x
e
l
-
b
a
se
d
a
d
a
p
t
i
v
e
seg
me
n
t
e
r
,
”
i
n
2
0
1
2
I
EEE
C
o
m
p
u
t
e
r
S
o
c
i
e
t
y
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
p
u
t
e
r
V
i
si
o
n
a
n
d
P
a
t
t
e
r
n
Re
c
o
g
n
i
t
i
o
n
Wo
r
k
sh
o
p
s
,
J
u
n
.
2
0
1
2
,
p
p
.
3
8
–
4
3
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
W
.
2
0
1
2
.
6
2
3
8
9
2
5
.
[
1
4
]
A
.
B
.
C
h
a
n
,
V
.
M
a
h
a
d
e
v
a
n
,
a
n
d
N
.
V
a
sco
n
c
e
l
o
s,
“
G
e
n
e
r
a
l
i
z
e
d
st
a
u
f
f
e
r
–
g
r
i
mso
n
b
a
c
k
g
r
o
u
n
d
s
u
b
t
r
a
c
t
i
o
n
f
o
r
d
y
n
a
mi
c
sce
n
e
s,”
Ma
c
h
i
n
e
V
i
si
o
n
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
2
,
n
o
.
5
,
p
p
.
7
5
1
–
7
6
6
,
S
e
p
.
2
0
1
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s0
0
1
3
8
-
0
1
0
-
0
2
6
2
-
3.
[
1
5
]
Z
.
Z
i
v
k
o
v
i
c
,
“
I
mp
r
o
v
e
d
a
d
a
p
t
i
v
e
G
a
u
ssi
a
n
mi
x
t
u
r
e
mo
d
e
l
f
o
r
b
a
c
k
g
r
o
u
n
d
su
b
t
r
a
c
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
1
7
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
P
a
t
t
e
r
n
R
e
c
o
g
n
i
t
i
o
n
,
2
0
0
4
.
I
C
P
R
2
0
0
4
.
,
2
0
0
4
,
p
p
.
2
8
-
3
1
V
o
l
.
2
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
P
R
.
2
0
0
4
.
1
3
3
3
9
9
2
.
[
1
6
]
“
T
M
S
3
2
0
C
6
6
7
8
mu
l
t
i
c
o
r
e
f
i
x
e
d
a
n
d
f
l
o
a
t
i
n
g
-
p
o
i
n
t
d
i
g
i
t
a
l
si
g
n
a
l
p
r
o
c
e
sso
r
,
”
T
e
x
a
s
I
n
st
r
u
me
n
t
s
2
0
1
4
.
A
c
c
e
sse
d
Ju
n
.
2
8
,
2
0
2
0
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
w
w
w
.
t
i
.
c
o
m/
l
i
t
/
d
s
/
sy
ml
i
n
k
/
t
ms
3
2
0
c
6
6
7
8
.
p
d
f
.
[
1
7
]
G
.
S
z
w
o
c
h
,
D
.
El
l
w
a
r
t
,
a
n
d
A
.
C
z
y
ż
e
w
sk
i
,
“
P
a
r
a
l
l
e
l
i
mp
l
e
m
e
n
t
a
t
i
o
n
o
f
b
a
c
k
g
r
o
u
n
d
su
b
t
r
a
c
t
i
o
n
a
l
g
o
r
i
t
h
ms
f
o
r
r
e
a
l
-
t
i
me
v
i
d
e
o
p
r
o
c
e
ssi
n
g
o
n
a
s
u
p
e
r
c
o
mp
u
t
e
r
p
l
a
t
f
o
r
m,”
J
o
u
r
n
a
l
o
f
R
e
a
l
-
T
i
m
e
I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
v
o
l
.
1
1
,
n
o
.
1
,
p
p
.
1
1
1
–
1
2
5
,
Ja
n
.
2
0
1
6
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
5
5
4
-
0
1
2
-
0
3
1
0
-
5.
[
1
8
]
L
.
M
a
b
r
o
u
k
,
S
.
H
u
e
t
,
S
.
B
e
l
k
o
u
c
h
,
D
.
H
o
u
z
e
t
,
Y
.
Z
e
n
n
a
y
i
,
a
n
d
A
.
H
a
mz
a
o
u
i
,
“
P
e
r
f
o
r
man
c
e
a
n
d
sc
a
l
a
b
i
l
i
t
y
i
mp
r
o
v
e
me
n
t
o
f
G
M
M
b
a
c
k
g
r
o
u
n
d
se
g
me
n
t
a
t
i
o
n
a
l
g
o
r
i
t
h
m
o
n
m
u
l
t
i
-
c
o
r
e
p
a
r
a
l
l
e
l
p
l
a
t
f
o
r
ms,”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
1
st
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
El
e
c
t
ro
n
i
c
En
g
i
n
e
e
ri
n
g
a
n
d
Re
n
e
w
a
b
l
e
En
e
r
g
y
(
I
C
EER
E
2
0
1
8
)
,
2
0
1
9
,
p
p
.
1
2
0
–
1
2
7
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
13
-
1
4
0
5
-
6
_
1
6
.
[
1
9
]
S
.
B
a
r
i
k
o
,
A
.
A
r
sal
a
n
e
,
A
.
K
l
i
l
o
u
,
a
n
d
A
.
A
b
o
u
n
a
d
a
,
“
Ef
f
i
c
i
e
n
t
p
a
r
a
l
l
e
l
i
mp
l
e
me
n
t
a
t
i
o
n
o
f
G
a
u
ssi
a
n
M
i
x
t
u
r
e
mo
d
e
l
b
a
c
k
g
r
o
u
n
d
su
b
t
r
a
c
t
i
o
n
a
l
g
o
r
i
t
h
m
o
n
a
n
e
mb
e
d
d
e
d
mu
l
t
i
-
c
o
r
e
d
i
g
i
t
a
l
si
g
n
a
l
p
r
o
c
e
sso
r
,
”
C
o
m
p
u
t
e
rs
a
n
d
El
e
c
t
ri
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
1
1
0
,
p
.
1
0
8
8
2
7
,
S
e
p
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
e
l
e
c
e
n
g
.
2
0
2
3
.
1
0
8
8
2
7
.
[
2
0
]
H
.
Ji
n
,
M
.
F
r
u
mk
i
n
,
a
n
d
J.
Y
a
n
,
“
A
u
t
o
mat
i
c
g
e
n
e
r
a
t
i
o
n
o
f
O
p
e
n
M
P
d
i
r
e
c
t
i
v
e
s
a
n
d
i
t
s
a
p
p
l
i
c
a
t
i
o
n
t
o
c
o
mp
u
t
a
t
i
o
n
a
l
f
l
u
i
d
d
y
n
a
m
i
c
s
c
o
d
e
s,”
i
n
H
i
g
h
Pe
r
f
o
rm
a
n
c
e
C
o
m
p
u
t
i
n
g
.
I
S
H
P
C
2
0
0
0
,
2
0
0
0
,
p
p
.
4
4
0
–
4
5
6
,
d
o
i
:
1
0
.
1
0
0
7
/
3
-
5
4
0
-
3
9
9
9
9
-
2
_
4
2
.
[
2
1
]
N
.
F
r
i
e
d
man
a
n
d
S
.
R
u
sse
l
l
,
“
I
mag
e
se
g
me
n
t
a
t
i
o
n
i
n
v
i
d
e
o
se
q
u
e
n
c
e
s:
a
p
r
o
b
a
b
i
l
i
s
t
i
c
a
p
p
r
o
a
c
h
,
”
a
r
Xi
v
:
1
3
0
2
.
1
5
3
9
,
2
0
1
3
.
[
2
2
]
T
e
x
a
s
I
n
st
r
u
me
n
t
s,
“
T
M
D
S
EV
M
6
6
7
8
L
EV
M
t
e
c
h
n
i
c
a
l
r
e
f
e
r
e
n
c
e
man
u
a
l
v
e
r
si
o
n
2
.
0
1
,
”
A
d
v
a
n
t
e
c
h
,
2
0
1
2
.
A
c
c
e
sse
d
S
e
p
.
2
6
,
2
0
2
1
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
w
w
w
.
f
a
r
n
e
l
l
.
c
o
m/
d
a
t
a
s
h
e
e
t
s/
1
6
9
1
4
4
2
.
p
d
f
.
[
2
3
]
A
.
K
l
i
l
o
u
,
F
.
B
o
u
r
z
e
i
x
,
O
.
B
o
u
r
j
a
,
Y
.
Z
e
n
n
a
y
i
,
L
.
M
a
b
r
o
u
k
,
a
n
d
S
.
B
e
l
k
o
u
c
h
,
“
R
e
a
l
-
t
i
me
p
a
r
a
l
l
e
l
i
mp
l
e
me
n
t
a
t
i
o
n
o
f
r
o
a
d
t
r
a
f
f
i
c
r
a
d
a
r
v
i
d
e
o
p
r
o
c
e
ssi
n
g
a
l
g
o
r
i
t
h
ms
o
n
a
p
a
r
a
l
l
e
l
a
r
c
h
i
t
e
c
t
u
r
e
b
a
se
d
o
n
D
S
P
a
n
d
A
R
M
p
r
o
c
e
sso
r
s,”
i
n
2
0
1
5
1
5
t
h
I
n
t
e
rn
a
t
i
o
n
a
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
I
mp
r
o
ve
d
a
u
t
o
ma
ted
p
a
r
a
llel imp
leme
n
ta
tio
n
o
f
GMM b
a
ck
g
r
o
u
n
d
s
u
b
tr
a
ctio
n
…
(
S
ma
il B
a
r
iko
)
559
C
o
n
f
e
re
n
c
e
o
n
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s D
e
si
g
n
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
(
I
S
D
A)
,
D
e
c
.
2
0
1
5
,
p
p
.
1
8
3
–
1
8
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
S
D
A
.
2
0
1
5
.
7
4
8
9
2
2
2
.
[
2
4
]
A
.
A
r
sal
a
n
e
,
N
.
El
B
a
r
b
r
i
,
K
.
R
h
o
f
i
r
,
A
.
T
a
b
y
a
o
u
i
,
a
n
d
A
.
K
l
i
l
o
u
,
“
B
u
i
l
d
i
n
g
a
p
o
r
t
a
b
l
e
d
e
v
i
c
e
b
a
se
d
o
n
D
S
P
f
o
r
me
a
t
d
i
s
c
r
i
mi
n
a
t
i
o
n
,
”
i
n
2
0
1
6
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
E
n
g
i
n
e
e
ri
n
g
&
MIS
(
I
C
EM
I
S
)
,
S
e
p
.
2
0
1
6
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
EM
I
S
.
2
0
1
6
.
7
7
4
5
3
2
7
.
[
2
5
]
S
.
B
a
r
i
k
o
,
A
.
K
l
i
l
o
u
,
A
.
A
b
o
u
n
a
d
a
,
a
n
d
A
.
A
r
sal
a
n
e
,
“
Ef
f
i
c
i
e
n
t
i
mp
l
e
me
n
t
a
t
i
o
n
o
f
b
a
c
k
g
r
o
u
n
d
s
u
b
t
r
a
c
t
i
o
n
G
M
M
me
t
h
o
d
o
n
d
i
g
i
t
a
l
si
g
n
a
l
p
r
o
c
e
sso
r
,
”
i
n
2
0
2
0
I
n
t
e
rn
a
t
i
o
n
a
l
S
y
m
p
o
s
i
u
m
o
n
A
d
v
a
n
c
e
d
E
l
e
c
t
r
i
c
a
l
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
T
e
c
h
n
o
l
o
g
i
e
s
(
I
S
AE
C
T
)
,
N
o
v
.
2
0
2
0
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
S
A
EC
T
5
0
5
6
0
.
2
0
2
0
.
9
5
2
3
6
7
1
.
[
2
6
]
A
.
K
l
i
l
o
u
a
n
d
A
.
A
r
sal
a
n
e
,
“
P
a
r
a
l
l
e
l
i
m
p
l
e
me
n
t
a
t
i
o
n
o
f
p
u
l
se
c
o
mp
r
e
ssi
o
n
me
t
h
o
d
o
n
a
m
u
l
t
i
-
c
o
r
e
d
i
g
i
t
a
l
si
g
n
a
l
p
r
o
c
e
sso
r
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
E
l
e
c
t
r
i
c
a
l
a
n
d
C
o
m
p
u
t
e
r
En
g
i
n
e
e
r
i
n
g
(
I
J
E
C
E
)
,
v
o
l
.
1
0
,
n
o
.
6
,
pp
.
6
5
4
1
-
6
5
4
8
,
D
e
c
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
c
e
.
v
1
0
i
6
.
p
p
6
5
4
1
-
6
5
4
8
.
[
2
7
]
R
.
v
a
n
d
e
r
P
a
s,
E
.
S
t
o
t
z
e
r
,
a
n
d
C
.
T
e
r
b
o
v
e
n
,
U
si
n
g
O
p
e
n
MP
—
t
h
e
n
e
x
t
s
t
e
p
:
a
f
f
i
n
i
t
y
,
a
c
c
e
l
e
r
a
t
o
rs
,
t
a
sk
i
n
g
,
a
n
d
S
I
MD
,
M
I
T
P
r
e
ss,
2
0
1
7
.
[
2
8
]
R
.
C
h
a
n
d
r
a
,
R
.
M
e
n
o
n
,
L
.
D
a
g
u
m,
D
.
K
o
h
r
,
D
.
M
a
y
d
a
n
,
a
n
d
J.
M
c
D
o
n
a
l
d
,
P
a
r
a
l
l
e
l
p
r
o
g
ra
m
m
i
n
g
i
n
O
p
e
n
MP
,
M
o
r
g
a
n
K
a
u
f
man
n
,
2
0
0
0
.
[
2
9
]
“
C
h
a
n
g
e
d
e
t
e
c
t
i
o
n
.
n
e
t
(
C
D
N
ET
)
,
A
v
i
d
e
o
d
a
t
a
b
a
se
f
o
r
t
e
st
i
n
g
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
a
l
g
o
r
i
t
h
ms
,
”
C
h
a
n
g
e
d
e
t
e
c
t
i
o
n
.
n
e
t
,
2
0
1
4
,
A
c
c
e
sse
d
Ju
n
.
1
3
,
2
0
2
0
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
c
h
a
n
g
e
d
e
t
e
c
t
i
o
n
.
n
e
t
/
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
S
m
a
il
B
a
r
ik
o
w
a
s
b
o
rn
in
M
a
rr
a
k
e
c
h
,
M
o
ro
c
c
o
.
He
re
c
e
iv
e
d
th
e
e
n
g
in
e
e
r’s
d
e
g
re
e
in
e
lec
tro
n
ics
a
n
d
e
m
b
e
d
d
e
d
sy
st
e
m
s
f
ro
m
ENS
A
-
M
a
rra
k
e
c
h
in
2
0
1
1
.
S
in
c
e
2
0
1
1
,
h
e
w
o
rk
e
d
o
n
d
if
f
e
re
n
t
in
d
u
str
ials
f
ield
:
a
e
ro
n
a
u
ti
c
s,
se
m
i
-
c
o
n
d
u
c
to
r
a
n
d
a
u
to
m
o
ti
v
e
c
o
m
p
a
n
ies
li
k
e
Zo
d
iac
A
e
ro
s
p
a
c
e
,
S
T
M
icro
e
lec
tro
n
ics
,
A
lt
ra
n
,
L
e
a
r
c
o
rp
o
ra
ti
o
n
a
n
d
S
tellan
ti
s.
His
-
sc
ien
ti
f
ic
re
se
a
rc
h
in
tere
sts
in
c
l
u
d
e
re
a
l
-
ti
m
e
im
p
le
m
e
n
tatio
n
o
f
im
a
g
e
a
n
d
v
id
e
o
p
r
o
c
e
ss
in
g
in
p
a
ra
ll
e
l
a
rc
h
it
e
c
tu
re
p
latf
o
rm
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
a
rik
o
_
sm
a
il
@
y
a
h
o
o
.
f
r
.
Abd
e
ss
a
m
a
d
K
li
lo
u
wa
s
b
o
rn
in
M
a
rra
k
e
c
h
,
M
o
ro
c
c
o
,
in
1
9
8
7
.
He
re
c
e
iv
e
d
a
n
e
n
g
in
e
e
r’s
d
e
g
re
e
in
2
0
1
0
a
n
d
a
P
h
.
D
.
d
e
g
re
e
in
2
0
1
6
f
ro
m
th
e
Un
iv
e
rsit
y
o
f
Ca
d
y
Ay
y
a
d
,
M
a
rra
k
e
c
h
,
M
o
ro
c
c
o
.
S
in
c
e
2
0
1
7
,
h
e
is
a
f
u
ll
p
ro
f
e
ss
o
r
a
t
th
e
De
p
a
rt
m
e
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
in
t
h
e
F
a
c
u
lt
y
o
f
S
c
ien
c
e
s
a
n
d
T
e
c
h
n
o
lo
g
y
,
S
u
lt
a
n
M
o
u
lay
S
li
m
a
n
e
Un
iv
e
r
sit
y
,
Be
n
i
M
e
ll
a
l
,
M
o
ro
c
c
o
.
His
a
re
a
o
f
re
s
e
a
rc
h
is
f
o
c
u
se
d
o
n
p
a
ra
ll
e
l
a
n
d
re
a
l
ti
m
e
o
p
ti
m
iza
ti
o
n
o
f
sig
n
a
l
p
r
o
c
e
ss
in
g
a
lg
o
rit
h
m
s
o
n
m
u
lt
i
-
c
o
re
a
n
d
m
u
lt
i
-
p
ro
c
e
ss
o
rs
p
a
ra
ll
e
l
m
a
c
h
in
e
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
.
k
li
lo
u
@u
sm
s.m
a
.
Abd
e
lo
u
a
h
e
d
A
b
o
u
n
a
d
a
w
a
s
b
o
rn
in
S
i
d
i
S
li
m
a
n
e
,
M
o
r
o
c
c
o
.
He
re
c
e
iv
e
d
h
is
P
h
.
D
.
d
e
g
re
e
f
ro
m
C
a
d
i
Ay
y
a
d
U
n
iv
e
rsity
in
1
9
9
2
.
In
a
d
d
it
i
o
n
,
h
e
re
c
e
iv
e
d
th
e
HD
R
d
e
g
re
e
in
2
0
1
3
f
ro
m
S
u
lt
a
n
M
o
u
lay
S
li
m
a
n
e
Un
iv
e
rsit
y
.
His
-
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
a
d
v
a
n
c
e
d
m
e
th
o
d
s
f
o
r
n
u
m
e
rica
l
a
u
to
m
a
t
ic
a
n
d
re
n
e
w
a
b
le
e
n
e
rg
y
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
.
a
b
o
u
n
a
d
a
@g
m
a
il
.
c
o
m
.
As
sia
Ar
s
a
la
n
e
re
c
e
iv
e
d
a
n
e
n
g
in
e
e
r’s
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
f
ro
m
th
e
Na
ti
o
n
a
l
S
c
h
o
o
l
o
f
A
p
p
li
e
d
S
c
ien
c
e
s
o
f
Kh
o
u
rib
g
a
in
2
0
1
4
a
n
d
a
P
h
.
D
.
d
e
g
re
e
in
2
0
1
9
f
ro
m
th
e
Un
iv
e
rsity
o
f
Ha
s
sa
n
I,
S
e
tt
a
t,
M
o
ro
c
c
o
.
S
h
e
is
a
f
u
ll
p
r
o
f
e
ss
o
r
in
th
e
De
p
a
rtm
e
n
t
o
f
M
e
c
h
a
tro
n
ics
,
Hig
h
S
c
h
o
o
l
o
f
T
e
c
h
n
o
l
o
g
ies
,
S
u
lt
a
n
M
o
u
lay
S
li
m
a
n
e
Un
i
v
e
rsity
,
Be
n
i
M
e
ll
a
l
,
M
o
ro
c
c
o
.
He
r
a
re
a
o
f
re
se
a
rc
h
i
n
c
lu
d
e
s
a
rti
f
icia
l
in
telli
g
e
n
c
e
,
m
a
c
h
in
e
v
isio
n
,
im
a
g
e
p
ro
c
e
ss
in
g
a
n
d
e
m
b
e
d
d
e
d
sy
ste
m
s.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
rsa
lan
.
a
ss
ia@
g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.