I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
,
p
p
.
1
5
5
5
~1
568
I
SS
N:
2
502
-
4
7
52
,
DOI
: 1
0
.
1
1
5
9
1
/ijee
cs
.v
37.
i
3
.
p
p
1
5
5
5
-
1
5
6
8
1555
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
Co
mpa
ra
tive a
na
ly
sis
of edge de
tec
tion ima
g
e proces
sing
techniqu
es for
efficient
traff
i
c sig
n
a
l ma
na
g
ement sys
tem
Nira
nja
na
C
ha
nd
ra
s
ek
a
ra
B
ha
ra
t
hy
,
M
a
nju
S
hiv
a
nn
a
D
a
s
a
pp
a
na
v
a
r
,
Anne G
o
wda
A
leri
B
y
re
G
o
wda
,
J
ij
esh
J
is
ha
J
a
na
rdha
na
n
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
o
n
i
c
s a
n
d
C
o
mm
u
n
i
c
a
t
i
o
n
En
g
i
n
e
e
r
i
n
g
,
S
r
i
V
e
n
k
a
t
e
sh
w
a
r
a
C
o
l
l
e
g
e
o
f
En
g
i
n
e
e
r
i
n
g
,
B
a
n
g
a
l
o
r
e
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
ul
2
,
2
0
2
4
R
ev
is
ed
Oct
4
,
2
0
2
4
Acc
ep
ted
Oct
7
,
2
0
2
4
Traffic
m
o
n
it
o
rin
g
a
n
d
c
o
n
tr
o
l
p
o
se
sig
n
ifi
c
a
n
t
c
h
a
ll
e
n
g
e
s,
p
a
rt
icu
larly
i
n
m
e
tro
p
o
li
tan
a
re
a
s
wo
rld
wid
e
.
T
ra
d
it
io
n
a
l
m
e
th
o
d
s
su
c
h
a
s
ti
m
e
rs,
traffic
p
o
li
c
e
c
o
n
tro
l,
a
n
d
se
n
so
r
-
b
a
se
d
sy
ste
m
s
h
a
v
e
b
e
c
o
m
e
in
c
re
a
sin
g
l
y
in
e
ffe
c
ti
v
e
d
u
e
t
o
e
sc
a
latin
g
traff
ic
v
o
l
u
m
e
s.
Im
a
g
e
p
ro
c
e
ss
in
g
e
m
e
rg
e
s
a
s
a
p
ro
m
isin
g
tec
h
n
o
l
o
g
y
fo
r
e
n
h
a
n
c
in
g
traffic
c
o
n
t
ro
l
sy
ste
m
s.
T
h
i
s
re
se
a
r
c
h
a
ims
to
a
d
d
re
ss
th
e
in
e
fficie
n
c
ies
o
f
c
u
rre
n
t
traffic
m
a
n
a
g
e
m
e
n
t
sy
ste
m
s
(TM
S
)
b
y
c
o
n
d
u
c
ti
n
g
a
d
e
tailed
c
o
m
p
a
ra
ti
v
e
a
n
a
ly
sis
o
f
c
o
n
v
e
n
t
io
n
a
l
ima
g
e
p
ro
c
e
ss
in
g
tec
h
n
iq
u
e
s,
fo
c
u
si
n
g
o
n
e
d
g
e
-
b
a
se
d
m
e
th
o
d
s.
Util
izin
g
th
e
o
p
e
n
so
u
rc
e
c
o
m
p
u
ter
v
isio
n
(
Op
e
n
CV
)
Li
b
ra
ry
,
we
e
v
a
l
u
a
ted
v
a
rio
u
s
e
d
g
e
d
e
tec
ti
o
n
tec
h
n
i
q
u
e
s
b
a
se
d
o
n
q
u
a
li
tati
v
e
p
a
ra
m
e
ters
li
k
e
e
n
v
i
ro
n
m
e
n
ta
l
li
g
h
t
in
g
,
o
b
jec
t
d
e
tec
ti
o
n
,
a
n
d
siz
e
,
a
s
we
ll
a
s q
u
a
n
ti
tati
v
e
p
a
ra
m
e
ters
su
c
h
a
s
a
v
e
ra
g
e
traffic
d
e
n
sity
,
sh
a
rp
n
e
ss
ra
ti
o
(S
R),
a
b
r
u
p
tn
e
ss
,
se
n
siti
v
it
y
fa
c
to
rs
(S
F
),
p
ro
c
e
ss
in
g
t
ime
,
fra
m
e
p
ro
c
e
ss
in
g
ti
m
e
,
a
n
d
a
v
e
ra
g
e
fra
m
e
s
p
e
r
se
c
o
n
d
(F
P
S
).
T
h
e
stu
d
y
f
in
d
s
th
a
t
t
h
e
C
a
n
n
y
e
d
g
e
d
e
tec
ti
o
n
tec
h
n
iq
u
e
o
u
tp
e
rfo
rm
s
o
th
e
rs,
wit
h
a
n
a
v
e
ra
g
e
traffic
d
e
n
sity
o
f
3
0
.
1
2
a
c
ro
ss
1
0
fra
m
e
s,
su
p
e
ri
o
r
S
R,
a
n
d
m
in
ima
l
p
r
o
c
e
ss
in
g
ti
m
e
o
f
9
9
se
c
o
n
d
s.
T
h
is
m
a
k
e
s
it
h
ig
h
ly
e
ffe
c
ti
v
e
fo
r
traffic
c
o
n
tr
o
l
a
p
p
li
c
a
ti
o
n
s
b
y
g
e
n
e
ra
ti
n
g
h
i
g
h
-
q
u
a
li
ty
e
d
g
e
m
a
p
s
with
m
in
ima
l
n
o
ise
.
Ou
r
fi
n
d
i
n
g
s
su
g
g
e
st
t
h
a
t
imp
ro
v
i
n
g
c
o
n
v
e
n
t
io
n
a
l
ima
g
e
p
ro
c
e
ss
in
g
tec
h
n
i
q
u
e
s
a
n
d
in
teg
ra
ti
n
g
d
e
e
p
lea
rn
in
g
a
l
g
o
rit
h
m
s
c
a
n
fu
rth
e
r
e
n
h
a
n
c
e
TM
S
,
lea
d
in
g
to
m
o
re
e
fficie
n
t
u
rb
a
n
m
o
b
il
i
ty
a
n
d
re
d
u
c
e
d
e
n
v
iro
n
m
e
n
tal
imp
a
c
t.
K
ey
w
o
r
d
s
:
C
an
n
y
ed
g
e
d
etec
tio
n
C
o
m
p
u
ter
v
is
io
n
lib
r
ar
y
I
m
ag
e
p
r
o
ce
s
s
in
g
I
m
ag
e
s
eg
m
en
tatio
n
Ma
ch
in
e
lear
n
in
g
Op
en
s
o
u
r
ce
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
:
Nir
an
jan
a
C
h
an
d
r
asek
ar
a
B
h
ar
ath
y
Dep
ar
tm
en
t o
f
E
lectr
o
n
ics an
d
C
o
m
m
u
n
icatio
n
E
n
g
in
ee
r
i
n
g
,
Sri
Ven
k
atesh
war
a
C
o
lleg
e
o
f
E
n
g
in
ee
r
i
n
g
B
an
g
alo
r
e,
Kar
n
atak
a
E
m
ail: n
ir
an
jan
a3
1
9
4
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
Me
tr
o
cities
all
o
v
er
th
e
g
lo
b
e
ar
e
s
u
f
f
er
in
g
f
r
o
m
s
ev
er
e
tr
af
f
ic
is
s
u
es.
T
h
e
tr
af
f
ic
jam
s
an
d
tr
af
f
ic
co
n
g
esti
o
n
r
esu
lt
in
in
c
r
ea
s
ed
tr
av
ellin
g
tim
e
f
o
r
s
h
o
r
t
d
is
tan
ce
s
,
ex
ce
s
s
f
u
el
co
n
s
u
m
p
tio
n
,
in
f
r
astru
ctu
r
e
s
tr
ain
,
n
eg
ativ
e
im
p
ac
t
o
n
l
if
e
q
u
ality
,
n
e
g
a
tiv
e
im
p
ac
t
o
n
th
e
ec
o
n
o
m
y
,
an
d
i
n
cr
ea
s
ed
air
p
o
llu
tio
n
.
I
n
ad
d
itio
n
t
o
th
at
in
cr
ea
s
ed
a
cc
id
en
ts
an
d
f
r
u
s
tr
atio
n
am
o
n
g
co
m
m
u
ter
s
ar
e
th
e
c
o
m
m
o
n
im
p
ac
ts
o
f
tr
af
f
ic
is
s
u
es.
T
h
e
in
cr
ea
s
e
in
tr
af
f
i
c
jam
s
an
d
co
n
g
esti
o
n
lead
s
r
ed
u
ctio
n
in
th
e
ef
f
icien
cy
o
f
th
e
tr
an
s
p
o
r
tatio
n
s
y
s
tem
.
No
wad
ay
s
,
tr
af
f
ic
m
an
ag
em
en
t
s
y
s
tem
s
(
T
MS)
[
1
]
in
clu
d
e
tr
af
f
ic
m
o
n
ito
r
in
g
a
n
d
d
ata
co
llectio
n
,
tr
af
f
ic
an
aly
s
is
an
d
p
r
ed
ictio
n
,
tr
af
f
ic
co
n
tr
o
l
an
d
r
e
g
u
latio
n
,
p
u
b
lic
in
f
o
r
m
atio
n
a
n
d
c
o
m
m
u
n
icatio
n
,
s
m
ar
t
in
f
r
astru
ctu
r
e
an
d
tech
n
o
lo
g
y
,
an
d
co
n
tin
u
o
u
s
m
o
n
ito
r
in
g
a
n
d
ev
alu
atio
n
.
T
MS
m
ajo
r
l
y
co
llect
r
ea
l
-
tim
e
d
ata
lik
e
tr
af
f
ic
d
en
s
ity
,
s
p
ee
d
,
c
o
n
g
esti
o
n
lev
el,
ty
p
e
o
f
v
eh
icle
[
2
]
,
an
d
wea
th
er
co
n
d
iti
o
n
s
an
d
b
ased
o
n
co
n
v
en
ie
n
t
tech
n
iq
u
es
lik
e
i
m
ag
e
p
r
o
ce
s
s
in
g
an
d
m
ac
h
in
e
lear
n
in
g
ap
p
r
o
ac
h
es,
tr
a
f
f
ic
p
atter
n
s
an
d
tr
en
d
s
will
b
e
id
en
tifie
d
.
T
h
e
c
o
n
tr
o
l
m
ea
s
u
r
es
ar
e
ta
k
en
b
ased
o
n
th
e
p
atter
n
an
d
tr
en
d
s
.
I
n
th
e
last
s
tag
e,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
555
-
1
5
6
8
1556
a
c
o
n
t
i
n
u
o
u
s
m
o
n
i
t
o
r
i
n
g
a
n
d
ev
a
l
u
a
t
i
o
n
p
r
o
c
e
s
s
is
c
a
r
r
ie
d
o
u
t
f
o
r
t
h
e
o
p
t
i
m
i
za
t
i
o
n
o
f
T
MS
.
I
m
a
g
e
p
r
o
c
e
s
s
i
n
g
[
3
]
is
th
e
m
o
s
t
p
o
p
u
la
r
tech
n
i
q
u
e
ad
o
p
ted
in
T
MS
d
u
e
to
e
n
h
an
ce
d
v
is
u
al
p
e
r
ce
p
tio
n
,
q
u
an
titativ
e
an
aly
s
is
,
m
o
n
ito
r
in
g
an
d
r
em
o
te
s
en
s
in
g
,
an
d
cr
ea
tiv
e
m
an
ip
u
latio
n
.
T
h
e
s
ig
n
if
ican
ce
o
f
im
ag
e
p
r
o
ce
s
s
in
g
is
en
ab
lin
g
m
o
n
ito
r
in
g
,
o
p
tim
is
atio
n
o
f
t
r
af
f
ic
f
lo
w,
d
ata
co
llec
ti
o
n
a
n
d
an
aly
s
is
[
4
]
.
T
h
e
im
p
o
r
ta
n
t
im
ag
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es u
s
ed
in
tr
af
f
ic
s
ig
n
al
m
an
ag
em
en
t a
r
e
p
r
esen
te
d
i
n
Fig
u
r
e
1
.
Fig
u
r
e
1
.
C
lass
if
icatio
n
o
f
im
a
g
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es f
o
r
tr
af
f
ic
s
ig
n
al
m
an
ag
e
m
en
t
I
m
ag
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
h
av
e
b
ec
o
m
e
a
n
in
teg
r
al
p
ar
t
o
f
tr
af
f
ic
m
an
a
g
em
en
t.
T
h
ey
o
f
f
er
b
en
ef
its
lik
e
r
ea
l
-
tim
e
m
o
n
it
o
r
in
g
,
an
aly
s
is
,
an
d
d
ec
is
io
n
-
m
ak
in
g
ca
p
ab
ilit
ies
at
d
if
f
e
r
en
t
lev
els
o
f
th
e
tr
an
s
p
o
r
tatio
n
n
etwo
r
k
.
T
h
e
b
l
o
ck
d
iag
r
am
o
f
im
ag
e
p
r
o
ce
s
s
in
g
in
tr
a
f
f
ic
s
ig
n
allin
g
is
s
h
o
w
n
in
Fig
u
r
e
2
.
Fig
u
r
e
2
.
I
m
ag
e
p
r
o
ce
s
s
in
g
in
tr
af
f
ic
s
ig
n
allin
g
s
y
s
tem
I
n
a
tr
af
f
ic
s
ig
n
al
m
an
ag
em
en
t
s
y
s
tem
illu
s
tr
ated
in
Fig
u
r
e
2
,
im
ag
e
p
r
o
ce
s
s
in
g
p
lay
s
a
c
r
u
cial
r
o
le
in
en
s
u
r
in
g
ef
f
icien
t
tr
af
f
ic
f
lo
w
an
d
s
af
ety
.
C
am
er
as
in
s
talled
at
in
ter
s
ec
ti
o
n
s
ca
p
tu
r
e
r
ea
l
-
tim
e
im
ag
es
o
f
tr
af
f
ic
co
n
d
itio
n
s
.
T
h
e
ca
p
tu
r
ed
im
ag
es
ar
e
s
eg
m
e
n
ted
t
o
id
en
tify
d
if
f
er
e
n
t
elem
en
ts
s
u
ch
as
v
eh
icles,
p
ed
estrian
s
,
an
d
b
ac
k
g
r
o
u
n
d
.
Usi
n
g
tech
n
iq
u
es
lik
e
o
b
ject
d
etec
tio
n
,
v
eh
icles
an
d
o
th
er
r
elev
an
t
o
b
jects
ar
e
d
etec
ted
wit
h
in
th
e
s
eg
m
en
ted
im
ag
es.
On
ce
d
etec
ted
,
th
e
s
y
s
tem
r
ec
o
g
n
izes
th
e
ty
p
e
o
f
v
eh
icles
(
ca
r
,
tr
u
ck
,
an
d
b
icy
cle
)
an
d
th
e
p
r
esen
ce
o
f
p
ed
estrian
s
.
T
h
e
s
y
s
tem
co
m
p
ar
es
th
e
cu
r
r
en
t
im
ag
e
d
ata
with
d
ata
f
r
o
m
p
r
ev
io
u
s
in
s
tan
ce
s
to
id
en
tif
y
ch
an
g
es
in
tr
af
f
ic
f
lo
w,
v
eh
icle
d
en
s
ity
,
o
r
o
th
er
r
e
lev
an
t
p
ar
am
eter
s
.
T
o
co
m
p
ar
e
th
e
tech
n
i
q
u
es
o
n
th
e
Op
en
C
V
p
latf
o
r
m
,
Var
io
u
s
p
er
f
o
r
m
a
n
ce
p
ar
am
eter
s
ca
n
b
e
tak
en
in
to
ac
co
u
n
t
.
T
h
e
p
er
f
o
r
m
an
c
e
p
ar
am
eter
s
s
u
ch
as
ac
cu
r
ac
y
,
tr
af
f
ic
d
en
s
ity
,
s
p
ee
d
,
r
o
b
u
s
tn
ess
,
ed
g
e
lo
ca
lizatio
n
,
m
em
o
r
y
u
s
ag
e,
co
lo
u
r
,
s
ize
o
f
o
b
jects
an
d
ea
s
in
ess
o
f
i
m
p
lem
en
tatio
n
[
5
]
ar
e
ta
k
en
in
to
c
o
n
s
id
er
atio
n
.
B
y
co
n
d
u
ctin
g
s
u
ch
co
m
p
ar
is
o
n
s
,
o
n
e
ca
n
d
eter
m
in
e
th
e
s
tr
en
g
th
s
an
d
wea
k
n
ess
es
o
f
ea
ch
tech
n
iq
u
e
an
d
ch
o
o
s
e
th
e
m
o
s
t su
itab
le
o
n
e
a
cc
o
r
d
in
g
to
th
e
a
p
p
licatio
n
’
s
s
p
ec
if
ic
r
eq
u
ir
e
m
en
ts
.
2.
L
I
T
E
R
AT
U
RE
SU
RVE
Y
I
m
ag
e
p
r
o
ce
s
s
in
g
is
o
n
e
o
f
th
e
m
o
s
t
p
r
o
m
is
in
g
tech
n
o
lo
g
ie
s
ty
p
ically
f
o
r
th
e
ap
p
licatio
n
s
lik
e
T
MS.
I
n
o
r
d
e
r
to
an
lay
s
e
th
e
s
u
itab
ilit
y
o
f
v
ar
io
u
s
im
ag
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es,
an
ex
ten
s
iv
e
l
iter
atu
r
e
r
ev
iew
is
ca
r
r
ied
o
u
t.
T
h
e
p
r
o
s
an
d
co
n
s
o
f
ea
ch
m
eth
o
d
ar
e
d
is
cu
s
s
ed
u
s
in
g
v
ar
io
u
s
liter
atu
r
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
C
o
mp
a
r
a
tive
a
n
a
lysi
s
o
f e
d
g
e
d
etec
tio
n
ima
g
e
p
r
o
ce
s
s
in
g
… (
N
ir
a
n
ja
n
a
C
h
a
n
d
r
a
s
ek
a
r
a
B
h
a
r
a
th
y
)
1557
2
.
1
.
Co
nv
ent
i
o
na
l t
ra
f
f
ic
s
ig
na
l m
a
na
g
e
m
ent
s
y
s
t
em
s
R
af
ter
et
a
l.
[
6
]
e
x
p
lain
s
th
e
f
ac
t
o
n
co
n
v
en
tio
n
al
tr
af
f
ic
s
ig
n
al
s
y
s
tem
s
h
av
e
lo
n
g
b
ee
n
th
e
co
r
n
er
s
to
n
e
o
f
u
r
b
an
tr
af
f
ic
m
an
a
g
em
en
t,
o
f
f
er
in
g
a
s
y
s
tem
atic
ap
p
r
o
ac
h
to
r
e
g
u
lat
in
g
v
e
h
icu
lar
an
d
p
ed
estrian
f
lo
w
at
in
ter
s
ec
tio
n
s
.
T
y
p
ically
,
th
ese
s
y
s
tem
s
r
el
y
o
n
p
r
e
-
p
r
o
g
r
am
m
ed
tim
i
n
g
s
eq
u
en
ce
s
th
at
d
ictate
wh
en
ea
ch
d
i
r
ec
tio
n
o
f
tr
af
f
ic
r
ec
eiv
es
a
g
r
ee
n
lig
h
t,
y
ield
in
g
to
th
e
n
ex
t
p
h
ase
o
f
m
o
v
em
e
n
t.
C
o
o
r
d
in
ated
with
s
en
s
o
r
s
o
r
t
im
er
s
,
th
ese
s
ig
n
als
en
s
u
r
e
ef
f
icien
t
tr
af
f
ic
p
r
o
g
r
ess
io
n
wh
i
le
also
p
r
io
r
itizin
g
s
af
ety
.
R
ed
u
cin
g
th
e
p
o
ten
tia
l
f
o
r
ac
ci
d
en
ts
an
d
co
n
g
esti
o
n
,
co
n
v
en
tio
n
al
tr
af
f
ic
s
ig
n
al
s
y
s
tem
s
ar
e
o
f
ten
au
g
m
en
ted
with
ad
d
itio
n
al
f
ea
tu
r
es.
J
ian
g
an
d
L
u
o
[
7
]
d
is
cu
s
s
ab
o
u
t
th
e
f
ac
t
th
at
co
n
v
en
tio
n
al
tr
af
f
ic
s
ig
n
al
s
y
s
tem
s
f
ac
e
ch
allen
g
es
in
d
y
n
am
ically
ad
a
p
tin
g
to
ch
a
n
g
in
g
tr
af
f
ic
p
atter
n
s
an
d
o
p
tim
izin
g
ef
f
icien
cy
in
r
ea
l
-
t
im
e.
Fix
ed
tim
in
g
s
eq
u
en
ce
s
m
ay
n
o
t
ad
eq
u
ately
ac
co
m
m
o
d
ate
f
lu
ctu
atio
n
s
in
tr
af
f
ic
v
o
lu
m
e,
lead
i
n
g
t
o
in
ef
f
icien
cies
an
d
p
o
ten
tial
co
n
g
esti
o
n
d
u
r
in
g
p
ea
k
h
o
u
r
s
.
Ad
d
itio
n
ally
,
th
es
e
s
y
s
tem
s
ca
n
s
tr
u
g
g
le
to
p
r
io
r
itize
alter
n
ativ
e
m
o
d
es
o
f
t
r
an
s
p
o
r
tatio
n
,
s
u
c
h
as
b
icy
cles
o
r
p
u
b
lic
tr
a
n
s
it,
wh
ich
ar
e
in
c
r
ea
s
in
g
ly
in
te
g
r
al
to
s
u
s
tain
ab
le
u
r
b
an
m
o
b
ilit
y
in
itiativ
es.
As
u
r
b
an
ce
n
ter
s
ev
o
lv
e
,
th
er
e
is
a
g
r
o
win
g
n
ee
d
f
o
r
t
r
af
f
ic
m
a
n
ag
em
en
t
s
o
l
u
tio
n
s
th
at
em
b
r
ac
e
tech
n
o
lo
g
ical
a
d
v
an
ce
m
e
n
ts
to
en
h
an
ce
a
d
a
p
tab
ilit
y
,
r
esp
o
n
s
iv
en
ess
,
an
d
in
clu
s
iv
ity
with
in
tr
an
s
p
o
r
tatio
n
n
etwo
r
k
s
.
Acc
o
r
d
in
g
to
K
u
m
ar
an
d
Sin
g
h
[
8
]
o
n
e
n
o
ta
b
le
d
is
ad
v
a
n
t
ag
e
o
f
co
n
v
en
tio
n
al
TMS
lie
s
in
th
eir
lim
ited
ad
ap
tab
ilit
y
to
d
y
n
a
m
ic
tr
af
f
ic
co
n
d
itio
n
s
.
Fix
ed
-
tim
in
g
s
ig
n
al
s
y
s
tem
s
,
f
o
r
in
s
tan
ce
,
o
p
er
ate
o
n
p
r
ed
eter
m
in
e
d
s
ch
ed
u
les
th
at
m
ay
n
o
t
ef
f
ec
tiv
ely
r
esp
o
n
d
to
f
lu
ctu
atio
n
s
in
tr
af
f
ic
v
o
l
u
m
e
th
r
o
u
g
h
o
u
t
th
e
d
ay
.
T
h
is
r
ig
id
ity
ca
n
lead
t
o
in
ef
f
icien
cies
an
d
co
n
g
esti
o
n
in
ar
ea
s
with
u
n
p
r
ed
ictab
le
tr
af
f
ic
p
atter
n
s
.
Mo
r
eo
v
er
,
co
n
v
en
tio
n
al
s
y
s
tem
s
o
f
ten
p
r
io
r
itize
v
eh
icu
la
r
tr
af
f
ic
o
v
er
alter
n
ativ
e
m
o
d
es
o
f
tr
an
s
p
o
r
tatio
n
,
s
u
ch
as
b
icy
cles
o
r
p
u
b
lic
tr
a
n
s
it,
n
eg
lectin
g
th
e
d
iv
er
s
e
n
ee
d
s
o
f
u
r
b
an
c
o
m
m
u
te
r
s
.
As
cities
ev
o
lv
e
an
d
tr
an
s
p
o
r
tatio
n
tr
en
d
s
s
h
if
t
to
w
ar
d
s
s
u
s
tain
ab
ilit
y
an
d
m
u
lti
-
m
o
d
al
s
o
lu
tio
n
s
,
th
er
e
’
s
a
p
r
e
s
s
in
g
n
ee
d
f
o
r
TMS
th
at
o
f
f
er
g
r
ea
ter
f
lex
ib
ilit
y
,
r
e
s
p
o
n
s
iv
en
ess
,
an
d
in
clu
s
iv
ity
.
2
.
2
.
I
ma
g
e
pro
ce
s
s
ing
us
ing
edg
e
det
ec
t
io
n
An
g
et
a
l.
[
9
]
h
ig
h
lig
h
ted
th
a
t
ed
g
e
d
etec
tio
n
is
ess
en
tial
f
o
r
v
ar
io
u
s
co
m
p
u
ter
v
is
io
n
ap
p
licatio
n
s
,
in
clu
d
in
g
o
b
ject
d
etec
tio
n
,
i
m
ag
e
s
eg
m
en
tatio
n
,
an
d
f
ea
t
u
r
e
ex
tr
ac
tio
n
.
T
h
e
p
r
im
a
r
y
o
b
jectiv
e
o
f
ed
g
e
d
etec
tio
n
is
to
lo
ca
te
s
ig
n
if
ican
t
tr
an
s
itio
n
s
o
r
d
i
s
co
n
tin
u
itie
s
in
p
ix
el
in
ten
s
ity
,
wh
ich
o
f
te
n
co
r
r
esp
o
n
d
to
th
e
ed
g
es
o
f
o
b
jects
o
r
ch
a
n
g
es
in
tex
tu
r
e
with
in
th
e
im
ag
e.
Acc
o
r
d
in
g
to
L
a
u
et
a
l.
[
1
0
]
th
er
e
a
r
e
s
ev
er
al
alg
o
r
ith
m
s
an
d
m
eth
o
d
s
em
p
lo
y
ed
f
o
r
ed
g
e
d
etec
tio
n
,
ea
ch
with
its
s
tr
en
g
th
s
an
d
li
m
itati
o
n
s
.
T
h
e
m
o
s
t
wid
ely
u
s
ed
tech
n
iq
u
e
is
th
e
g
r
ad
ien
t
-
b
ased
ap
p
r
o
ac
h
,
wh
e
r
e
g
r
ad
ie
n
ts
in
p
ix
el
in
ten
s
ity
ar
e
co
m
p
u
ted
u
s
in
g
o
p
er
ato
r
s
lik
e
th
e
So
b
el,
Pre
witt,
o
r
R
o
b
er
ts
o
p
er
ato
r
s
.
T
h
ese
o
p
er
ato
r
s
an
aly
ze
th
e
in
ten
s
ity
v
ar
iatio
n
in
th
e
n
eig
h
b
o
r
h
o
o
d
o
f
ea
ch
p
ix
el
to
id
en
tif
y
r
e
g
io
n
s
o
f
r
ap
id
ch
an
g
e,
i
n
d
ic
atin
g
p
o
te
n
tial
ed
g
es.
Wu
et
a
l.
[
1
1
]
r
ec
o
m
m
en
d
ed
th
at
E
d
g
e
d
ete
ctio
n
in
im
ag
e
p
r
o
ce
s
s
in
g
o
f
f
er
s
s
ev
er
al
ad
v
an
tag
es,
in
clu
d
in
g
its
ab
ilit
y
to
ex
tr
ac
t
im
p
o
r
tan
t
f
ea
tu
r
es
s
u
c
h
as
o
b
ject
b
o
u
n
d
ar
ies
an
d
c
o
n
to
u
r
s
,
w
h
ich
a
r
e
ess
en
tials
f
o
r
task
s
lik
e
o
b
ject
r
ec
o
g
n
itio
n
,
im
ag
e
s
eg
m
e
n
tatio
n
,
an
d
s
h
ap
e
an
al
y
s
is
.
B
y
h
ig
h
lig
h
tin
g
a
r
ea
s
o
f
r
a
p
id
in
te
n
s
ity
ch
an
g
e,
e
d
g
e
d
etec
tio
n
alg
o
r
ith
m
s
ca
n
ef
f
ec
tiv
ely
r
ed
u
ce
th
e
co
m
p
lex
ity
o
f
im
ag
es
wh
ile
p
r
eser
v
in
g
es
s
en
tial
in
f
o
r
m
atio
n
,
lead
in
g
to
m
o
r
e
ef
f
icien
t
p
r
o
c
ess
in
g
an
d
an
aly
s
is
.
Fu
r
th
er
m
o
r
e,
ed
g
e
d
etec
tio
n
f
ac
ilit
ates
th
e
id
en
tific
atio
n
o
f
k
ey
s
tr
u
ctu
r
es
with
in
im
ag
es,
aid
in
g
in
th
e
in
ter
p
r
etatio
n
a
n
d
u
n
d
er
s
tan
d
i
n
g
o
f
v
is
u
al
d
a
ta
in
v
ar
i
o
u
s
f
ield
s
s
u
ch
as m
ed
ical
im
ag
in
g
,
r
em
o
t
e
s
en
s
in
g
,
an
d
r
o
b
o
tics
.
Mo
s
av
i
et
a
l.
[
1
2
]
d
is
cu
s
s
es
ab
o
u
t
th
e
p
r
o
s
an
d
co
n
s
o
f
ed
g
e
d
etec
tio
n
tech
n
iq
u
e
i
n
d
etails.
On
e
s
ig
n
if
ican
t
d
is
ad
v
an
tag
e
is
it
s
s
en
s
i
tiv
ity
to
n
o
is
e,
wh
i
ch
ca
n
lead
to
th
e
d
etec
tio
n
o
f
f
alse
ed
g
es
an
d
in
ac
cu
r
ac
ies,
esp
ec
ially
in
im
ag
es
with
lo
w
s
ig
n
al
-
to
-
n
o
is
e
r
atio
s
.
E
d
g
e
d
etec
tio
n
alg
o
r
i
th
m
s
m
ay
s
tr
u
g
g
le
with
v
ar
iatio
n
s
in
lig
h
tin
g
co
n
d
itio
n
s
,
as
well
as
w
ith
th
e
p
r
esen
ce
o
f
tex
tu
r
e
o
r
p
atter
n
s
th
at
ca
n
in
ter
f
er
e
with
ed
g
e
id
en
tific
atio
n
.
T
h
e
ch
o
ice
o
f
p
ar
am
eter
s
a
n
d
alg
o
r
ith
m
s
f
o
r
ed
g
e
d
etec
tio
n
ca
n
s
ig
n
if
ican
tly
im
p
ac
t
th
e
r
esu
lts
,
r
eq
u
ir
in
g
ca
r
ef
u
l
tu
n
in
g
an
d
ex
p
e
r
im
en
tatio
n
t
o
ac
h
iev
e
o
p
tim
a
l
p
er
f
o
r
m
an
ce
ac
r
o
s
s
d
if
f
er
en
t
ty
p
es
o
f
im
ag
es
an
d
ap
p
licatio
n
s
.
E
d
g
e
d
etec
tio
n
r
em
ain
s
a
v
alu
ab
le
to
o
l
in
im
ag
e
p
r
o
ce
s
s
in
g
,
o
f
f
er
in
g
in
d
is
p
en
s
ab
le
in
s
ig
h
ts
in
to
th
e
s
tr
u
ctu
r
e
an
d
c
o
n
ten
t
o
f
v
is
u
al
d
ata.
T
h
is
m
eth
o
d
ef
f
ec
ti
v
ely
r
ed
u
ce
s
n
o
is
e
wh
ile
ac
cu
r
ate
ly
d
etec
tin
g
ed
g
es
with
h
ig
h
lo
ca
lizatio
n
an
d
m
in
im
al
f
alse
p
o
s
itiv
es.
Ho
wev
er
,
e
d
g
e
d
etec
tio
n
alg
o
r
ith
m
s
m
u
s
t c
o
n
ten
d
with
ch
allen
g
es su
ch
as n
o
is
e,
v
ar
y
in
g
lig
h
tin
g
co
n
d
itio
n
s
,
an
d
th
e
p
r
esen
ce
o
f
wea
k
o
r
s
p
u
r
io
u
s
ed
g
es,
wh
ich
ca
n
im
p
ac
t
th
eir
r
eliab
ilit
y
an
d
r
o
b
u
s
tn
ess
in
p
r
ac
tical
ap
p
licatio
n
s
.
No
n
eth
eless
,
with
o
n
g
o
in
g
ad
v
an
ce
m
en
ts
in
co
m
p
u
ter
v
is
io
n
an
d
im
ag
e
p
r
o
ce
s
s
in
g
,
e
d
g
e
d
etec
tio
n
r
e
m
ain
s
a
cr
itical
to
o
l
f
o
r
e
x
t
r
a
c
t
i
n
g
m
ea
n
i
n
g
f
u
l
i
n
f
o
r
m
a
t
i
o
n
a
n
d
p
o
we
r
i
n
g
a
w
i
d
e
r
a
n
g
e
o
f
a
p
p
l
ic
a
ti
o
n
s
a
c
r
o
s
s
d
i
v
e
r
s
e
d
o
m
a
i
n
s
[
1
3
]
.
2
.
3
.
I
ma
g
e
pro
ce
s
s
ing
us
ing
s
eg
m
ent
a
t
io
n
Sad
o
wsk
i
et
a
l.
[
1
4
]
s
tated
th
at
im
ag
e
p
r
o
ce
s
s
in
g
u
s
in
g
s
eg
m
en
tatio
n
is
v
ital
f
o
r
a
r
an
g
e
o
f
ap
p
licatio
n
s
,
in
clu
d
in
g
o
b
ject
d
etec
tio
n
,
m
e
d
ical
im
ag
in
g
,
a
n
d
s
atellite
im
ag
e
a
n
aly
s
is
.
B
y
d
iv
id
i
n
g
a
n
im
a
g
e
in
to
m
ea
n
in
g
f
u
l
s
eg
m
en
ts
,
s
eg
m
en
tatio
n
en
ab
les
th
e
ex
tr
ac
ti
o
n
o
f
im
p
o
r
tan
t
f
ea
t
u
r
es
an
d
t
h
e
id
en
tific
atio
n
o
f
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
555
-
1
5
6
8
1558
o
b
jects
o
f
in
ter
est
with
g
r
ea
t
er
ac
cu
r
ac
y
an
d
ef
f
icie
n
cy
.
T
h
is
p
r
o
ce
s
s
allo
ws
f
o
r
b
etter
u
n
d
er
s
tan
d
i
n
g
an
d
in
ter
p
r
etatio
n
o
f
v
is
u
al
d
ata,
f
ac
ilit
atin
g
task
s
lik
e
im
ag
e
r
ec
o
g
n
itio
n
,
m
ea
s
u
r
em
e
n
t,
an
d
class
if
icatio
n
.
Sau
r
a
et
a
l
.
[
1
5
]
m
en
tio
n
e
d
th
at
Seg
m
en
tatio
n
m
eth
o
d
s
ca
n
b
e
b
r
o
ad
ly
d
i
v
id
ed
as
two
m
ain
ty
p
es:
th
r
esh
o
ld
-
b
ased
an
d
r
e
g
io
n
-
b
ased
.
T
h
r
es
h
o
ld
-
b
ased
s
eg
m
e
n
tatio
n
in
v
o
lv
es
s
ettin
g
th
r
esh
o
ld
s
o
n
p
ix
e
l
in
ten
s
ity
o
r
co
lo
r
to
s
ep
ar
ate
o
b
jects
f
r
o
m
th
e
b
ac
k
g
r
o
u
n
d
.
T
h
is
m
eth
o
d
is
s
i
m
p
le
an
d
ef
f
ec
tiv
e
f
o
r
i
m
ag
e
s
with
well
-
d
ef
in
ed
b
o
u
n
d
ar
ies
an
d
u
n
if
o
r
m
b
ac
k
g
r
o
u
n
d
.
On
t
h
e
o
th
er
h
an
d
,
r
e
g
io
n
-
b
ased
s
eg
m
en
tatio
n
tech
n
iq
u
es
g
r
o
u
p
p
ix
els
to
g
eth
er
b
ased
o
n
s
im
ilar
itie
s
in
th
eir
attr
ib
u
tes,
s
u
ch
as
co
lo
r
o
r
tex
tu
r
e,
u
s
in
g
alg
o
r
ith
m
s
lik
e
r
e
g
i
on
g
r
o
win
g
o
r
s
p
lit
-
an
d
-
m
er
g
e.
T
h
ese
m
eth
o
d
s
ar
e
m
o
r
e
v
er
s
atile
an
d
r
o
b
u
s
t,
ca
p
ab
le
o
f
h
an
d
lin
g
co
m
p
lex
im
ag
es
with
v
ar
y
in
g
in
ten
s
ity
lev
els
an
d
tex
tu
r
es.
Acc
o
r
d
i
n
g
to
Nallap
er
u
m
a
et
a
l
.
[
1
6
]
im
ag
e
p
r
o
ce
s
s
in
g
u
s
in
g
s
eg
m
en
tatio
n
also
p
r
e
s
en
ts
ce
r
tain
ch
alle
n
g
es,
s
u
ch
as
th
e
s
elec
tio
n
o
f
ap
p
r
o
p
r
iate
s
eg
m
en
tatio
n
alg
o
r
ith
m
s
an
d
p
ar
am
eter
s
,
wh
ich
ca
n
s
ig
n
if
ican
tly
im
p
a
ct
th
e
r
esu
lts
.
Mo
r
eo
v
er
,
s
eg
m
en
tatio
n
m
ay
b
e
in
f
lu
en
ce
d
b
y
f
a
cto
r
s
lik
e
n
o
is
e,
illu
m
in
atio
n
v
ar
iatio
n
s
,
a
n
d
o
b
ject
o
cc
lu
s
io
n
s
,
lea
d
in
g
to
i
n
ac
c
u
r
ac
ies
o
r
in
co
m
p
lete
s
eg
m
en
tatio
n
.
D
esp
ite
th
ese
ch
allen
g
es,
s
eg
m
en
tatio
n
r
em
ain
s
a
p
o
wer
f
u
l
to
o
l
in
im
ag
e
p
r
o
ce
s
s
in
g
,
o
f
f
er
in
g
v
alu
a
b
le
in
s
ig
h
ts
an
d
en
ab
li
n
g
a
wid
e
r
a
n
g
e
o
f
ap
p
licatio
n
s
ac
r
o
s
s
d
iv
er
s
e
f
ield
s
.
2
.
4
.
O
t
her
m
et
ho
ds
Sch
u
lz
et
a
l.
[
1
7
]
s
u
g
g
ested
th
at
m
ac
h
in
e
lear
n
in
g
p
lay
s
a
p
iv
o
tal
r
o
le
in
im
ag
e
p
r
o
ce
s
s
in
g
b
y
p
r
o
v
id
i
n
g
p
o
wer
f
u
l
to
o
ls
a
n
d
alg
o
r
ith
m
s
in
s
ak
e
o
f
task
s
s
u
ch
as
o
b
ject
d
etec
t
io
n
,
class
if
icatio
n
,
s
eg
m
en
tatio
n
,
a
n
d
e
n
h
an
ce
m
e
n
t.
T
h
r
o
u
g
h
th
e
u
tili
za
tio
n
o
f
lar
g
e
d
atasets
an
d
ad
v
a
n
ce
d
a
lg
o
r
ith
m
s
,
m
ac
h
in
e
lear
n
in
g
m
o
d
els
ca
n
ac
q
u
ir
e
in
tr
icate
p
atter
n
s
an
d
r
elatio
n
s
h
ip
s
with
in
im
ag
es,
en
ab
lin
g
th
em
to
m
ak
e
ac
cu
r
ate
p
r
ed
ictio
n
s
an
d
d
ec
is
io
n
s
.
I
n
im
a
g
e
class
if
icatio
n
,
f
o
r
e
x
am
p
le,
co
n
v
o
lu
tio
n
a
l
n
eu
r
al
n
etwo
r
k
s
(
C
NNs)
[
1
8
]
a
r
e
co
m
m
o
n
l
y
e
m
p
lo
y
ed
to
au
to
m
atica
lly
id
e
n
tify
an
d
ca
te
g
o
r
ize
o
b
jects
with
in
im
ag
es
b
ased
o
n
th
eir
f
ea
tu
r
es.
Similar
ly
,
in
o
b
ject
d
etec
tio
n
,
m
o
d
els
lik
e
f
aster
R
-
C
N
N
an
d
YOL
O
u
tili
ze
m
ac
h
in
e
lear
n
in
g
tec
h
n
iq
u
es
to
lo
ca
liz
e
an
d
class
if
y
o
b
jects
o
f
in
ter
est
with
in
im
ag
es
with
r
e
m
ar
k
ab
l
e
p
r
ec
is
io
n
an
d
s
p
ee
d
.
Mo
r
eo
v
er
,
m
ac
h
in
e
lea
r
n
in
g
alg
o
r
ith
m
s
en
a
b
le
s
o
p
h
i
s
ticated
im
ag
e
en
h
an
ce
m
e
n
t
t
ec
h
n
iq
u
es,
s
u
ch
as
s
u
p
er
-
r
eso
lu
tio
n
a
n
d
d
e
n
o
is
in
g
,
b
y
lear
n
in
g
f
r
o
m
e
x
am
p
les
to
g
en
er
ate
h
ig
h
-
q
u
ality
im
ag
es
f
r
o
m
lo
w
-
r
es
o
lu
tio
n
o
r
n
o
is
y
in
p
u
ts
.
Y
u
et
a
l.
[
1
9
]
u
s
ed
Q
-
lea
r
n
in
g
an
d
f
ast
g
r
ad
ien
t
-
d
escen
t
to
f
ig
u
r
e
o
u
t
th
e
b
est
ac
tio
n
f
o
r
s
ig
n
al
s
ettin
g
s
.
T
h
ey
s
h
o
wed
h
o
w
well
th
is
m
et
h
o
d
w
o
r
k
s
with
s
im
u
latio
n
s
.
T
h
ey
aim
e
d
to
cu
t
d
o
wn
waitin
g
tim
e
b
y
d
ec
id
in
g
o
n
s
ig
n
al
tim
in
g
u
s
in
g
a
lin
e
ar
r
eg
r
ess
io
n
alg
o
r
ith
m
.
Pre
d
ictin
g
tr
af
f
ic
v
o
lu
m
e
ac
cu
r
ately
is
k
ey
f
o
r
s
ch
ed
u
li
n
g
tr
af
f
ic
s
ig
n
als.
T
o
tac
k
le
th
is
,
an
o
th
er
g
r
o
u
p
s
u
g
g
ested
u
s
in
g
a
co
m
b
in
atio
n
o
f
m
ac
h
in
e
lear
n
in
g
m
o
d
els to
f
o
r
ec
ast tr
af
f
ic
o
n
in
c
o
m
in
g
r
o
ad
s
.
Hu
an
g
et
a
l.
[
2
0
]
s
u
g
g
ested
an
ad
ap
tiv
e
tr
af
f
ic
s
ig
n
al
co
n
tr
o
ller
u
s
in
g
lin
ea
r
r
e
g
r
ess
io
n
to
p
r
e
d
ict
g
r
ee
n
lig
h
t
d
u
r
atio
n
f
o
r
ea
ch
p
h
ase,
r
ed
u
cin
g
v
e
h
icle
q
u
eu
es
d
u
r
in
g
r
e
d
p
er
io
d
s
.
An
o
t
h
er
r
esear
ch
p
r
o
p
o
s
es
a
d
y
n
am
ic
s
ig
n
al
tim
in
g
s
tr
ateg
y
b
ased
o
n
p
e
d
estrian
v
o
lu
m
e,
u
s
in
g
r
ad
a
r
d
ata
to
m
in
im
i
ze
waitin
g
d
e
la
y
s
.
T
h
ey
u
s
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
to
esti
m
ate
q
u
eu
ein
g
d
elay
s
.
A
n
ew
tr
af
f
ic
s
ig
n
al
s
witch
in
g
s
y
s
tem
is
in
tr
o
d
u
ce
d
,
in
teg
r
atin
g
o
b
ject
d
etec
tio
n
th
r
o
u
g
h
m
ac
h
i
n
e
lear
n
in
g
.
An
ev
o
lu
ti
o
n
ar
y
alg
o
r
ith
m
o
p
tim
izes
s
ig
n
al
s
tr
ateg
ies
to
cu
t
d
o
wn
o
n
p
ed
estrian
an
d
v
eh
icle
wait
tim
es.
Xu
et
a
l.
[
21]
p
r
o
p
o
s
ed
a
n
ew
way
to
g
ath
er
im
p
o
r
tan
t
tr
af
f
ic
d
ata
l
ik
e
v
eh
icle
v
o
lu
m
e
an
d
r
ed
-
lig
h
t
d
u
r
atio
n
at
s
ig
n
alize
d
h
ig
h
way
s
.
T
h
ey
u
s
e
a
k
-
n
ea
r
est
n
eig
h
b
o
u
r
class
if
icatio
n
alg
o
r
ith
m
an
d
im
ag
e
p
r
o
ce
s
s
in
g
.
T
h
is
d
ata
h
elp
s
im
p
r
o
v
e
s
ig
n
al
s
ettin
g
s
.
Oth
er
r
esear
ch
er
s
in
[
2
2
]
,
[
2
3
]
u
s
e
f
u
zz
y
lo
g
ic,
ar
tific
ial
n
e
u
r
al
n
etwo
r
k
s
,
an
d
g
en
etic
alg
o
r
ith
m
s
to
o
p
tim
ize
tr
af
f
ic
s
ig
n
als.
W
o
r
ld
wid
e
,
s
c
ien
tis
ts
ar
e
co
n
tin
u
o
u
s
ly
im
p
r
o
v
in
g
s
ig
n
al
o
p
er
atio
n
s
at
i
n
ter
s
ec
tio
n
s
to
m
a
k
e
tr
af
f
ic
co
n
tr
o
ller
s
s
m
ar
ter
.
Ma
ch
in
e
lear
n
in
g
is
h
ea
v
ily
u
s
ed
f
o
r
th
is
p
u
r
p
o
s
e.
R
ec
en
t
s
tu
d
ies
in
[
2
4
]
,
[
2
5
]
em
p
lo
y
r
ei
n
f
o
r
ce
m
en
t le
ar
n
i
n
g
to
cr
ea
te
ad
a
p
tab
le
an
d
s
m
ar
t tr
af
f
ic
s
ig
n
al
co
n
tr
o
ller
s
.
3.
M
AT
E
R
I
AL
S AN
D
M
E
T
H
O
DS
I
m
ag
e
p
r
o
ce
s
s
in
g
is
v
ital
f
o
r
tr
af
f
ic
s
ig
n
al
m
a
n
ag
em
e
n
t
in
cities.
C
am
er
as
ca
p
t
u
r
e
im
ag
es
at
in
ter
s
ec
tio
n
s
to
d
etec
t
v
eh
icles
an
d
p
ed
estrian
s
,
an
d
r
ef
i
n
e
s
ig
n
al
tim
in
g
s
to
allev
iate
co
n
g
esti
o
n
an
d
en
h
a
n
ce
tr
af
f
ic
f
lo
w
.
T
h
is
tech
n
o
lo
g
y
is
cr
u
cial
f
o
r
im
p
r
o
v
in
g
tr
a
f
f
ic
ef
f
icien
cy
an
d
s
af
ety
o
n
b
u
s
y
u
r
b
an
r
o
ad
s
.
I
n
o
r
d
er
to
an
al
y
s
e
th
e
p
er
f
o
r
m
an
ce
o
f
co
n
v
en
ti
o
n
al
tech
n
iq
u
es,
p
o
p
u
lar
alg
o
r
ith
m
s
co
m
e
u
n
d
e
r
th
ese
m
eth
o
d
s
ar
e
a
n
aly
s
ed
an
d
test
ed
u
s
in
g
Op
e
n
C
V
en
v
ir
o
n
m
en
t.
3
.
1
.
E
dg
e
det
ec
t
io
n t
ec
hn
iqu
es
E
d
g
e
d
ete
ctio
n
tech
n
iq
u
e
in
i
m
ag
e
p
r
o
ce
s
s
in
g
is
a
m
eth
o
d
u
s
ed
to
id
en
tify
b
o
u
n
d
ar
ies
with
in
an
im
ag
e,
h
i
g
h
lig
h
tin
g
a
b
r
u
p
t
c
h
an
g
es
in
in
ten
s
ity
.
T
h
is
h
el
p
s
in
e
x
tr
ac
tin
g
im
p
o
r
tan
t
f
e
atu
r
es
lik
e
e
d
g
es
o
r
o
u
tlin
es,
f
ac
ilit
atin
g
f
u
r
th
er
a
n
aly
s
is
an
d
p
r
o
ce
s
s
in
g
.
E
d
g
e
d
e
tectio
n
in
im
ag
e
p
r
o
ce
s
s
in
g
is
ess
en
tial
f
o
r
id
en
tify
in
g
b
o
u
n
d
a
r
ies
with
in
an
im
a
g
e,
wh
ich
h
ig
h
lig
h
ts
ab
r
u
p
t
ch
an
g
es
in
in
ten
s
ity
.
T
h
is
f
ac
ilit
ates
th
e
ex
tr
ac
tio
n
o
f
cr
u
cial
f
ea
t
u
r
es su
ch
as e
d
g
es o
r
o
u
tlin
es,
aid
i
n
g
f
u
r
th
er
an
aly
s
is
an
d
p
r
o
ce
s
s
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
C
o
mp
a
r
a
tive
a
n
a
lysi
s
o
f e
d
g
e
d
etec
tio
n
ima
g
e
p
r
o
ce
s
s
in
g
… (
N
ir
a
n
ja
n
a
C
h
a
n
d
r
a
s
ek
a
r
a
B
h
a
r
a
th
y
)
1559
3
.
1
.
1
.
Ca
nn
y
edg
e
det
ec
t
io
n
C
an
n
y
ed
g
e
d
etec
tio
n
is
a
wid
ely
u
s
ed
im
ag
e
p
r
o
ce
s
s
in
g
tec
h
n
iq
u
e
k
n
o
wn
f
o
r
its
h
ig
h
ac
c
u
r
ac
y
an
d
lo
w
er
r
o
r
r
ates
in
ed
g
e
d
etec
tio
n
.
T
h
e
m
eth
o
d
in
v
o
lv
es
s
e
v
er
al
s
tep
s
,
in
clu
d
i
n
g
n
o
is
e
r
ed
u
ctio
n
,
g
r
a
d
ien
t
ca
lcu
latio
n
,
n
o
n
-
m
ax
im
u
m
s
u
p
p
r
ess
io
n
,
an
d
ed
g
e
tr
ac
k
in
g
b
y
h
y
s
ter
esis
,
wh
ich
to
g
eth
e
r
en
s
u
r
e
p
r
ec
is
e
ed
g
e
d
etec
tio
n
s
u
itab
le
f
o
r
ap
p
licatio
n
s
s
u
ch
as
o
b
ject
r
ec
o
g
n
itio
n
an
d
im
a
g
e
s
eg
m
e
n
tatio
n
.
T
h
e
f
lo
wc
h
ar
t
o
f
th
e
C
an
n
y
ed
g
e
d
etec
tio
n
tec
h
n
iq
u
e
is
illu
s
tr
ated
in
Fig
u
r
e
3
.
T
h
e
alg
o
r
ith
m
b
eg
i
n
s
b
y
ta
k
in
g
an
im
ag
e
as
in
p
u
t
an
d
p
r
o
ce
e
d
s
th
r
o
u
g
h
th
ese
s
t
ag
es
to
ef
f
ec
tiv
ely
id
en
tify
ed
g
es
wh
ile
m
in
im
izin
g
n
o
is
e.
T
h
is
im
ag
e
is
th
en
co
n
v
er
ted
to
g
r
a
y
s
ca
le,
s
im
p
lify
in
g
th
e
p
r
o
ce
s
s
in
g
.
Nex
t,
a
Gau
s
s
ian
b
lu
r
is
a
p
p
lied
t
o
th
e
g
r
ay
s
ca
le
im
a
g
e
t
o
r
ed
u
ce
an
y
n
o
is
e
p
r
esen
t
in
th
e
im
ag
e.
T
h
e
C
an
n
y
e
d
g
e
d
etec
tio
n
alg
o
r
ith
m
is
r
en
o
wn
ed
f
o
r
its
h
ig
h
ac
cu
r
ac
y
an
d
lo
w
er
r
o
r
r
ates.
I
t
in
v
o
lv
es
th
e
f
o
llo
win
g
s
tep
s
,
ea
ch
co
n
tr
ib
u
tin
g
to
its
ef
f
ec
t
iv
en
ess
in
v
a
r
io
u
s
ap
p
licatio
n
s
lik
e
o
b
ject
r
ec
o
g
n
itio
n
an
d
im
a
g
e
s
eg
m
en
tatio
n
:
a)
I
n
p
u
t
im
a
g
e:
s
tar
t
b
y
lo
a
d
in
g
a
n
im
ag
e
an
d
co
n
v
er
tin
g
it to
g
r
ay
s
ca
le
f
o
r
s
im
p
lific
atio
n
.
b)
No
is
e
r
ed
u
ctio
n
: a
p
p
ly
a
Gau
s
s
ian
b
lu
r
to
th
e
g
r
ay
s
ca
le
im
ag
e
to
r
ed
u
ce
n
o
is
e.
c)
Gr
ad
ien
t
ca
lcu
latio
n
:
c
o
m
p
u
t
e
th
e
ca
lcu
latio
n
at
ea
ch
p
ix
el,
h
ig
h
lig
h
tin
g
ar
ea
s
with
r
ap
id
in
ten
s
ity
ch
an
g
es.
d)
No
n
-
m
ax
im
u
m
s
u
p
p
r
ess
io
n
:
i
d
en
tify
lo
ca
l
m
a
x
im
a
i
n
th
e
g
r
ad
ien
t
d
ir
ec
tio
n
a
n
d
s
u
p
p
r
ess
n
o
n
-
m
a
x
im
u
m
p
ix
els,
p
r
eser
v
in
g
th
in
e
d
g
es.
e)
T
h
r
esh
o
ld
in
g
:
d
ef
in
e
h
ig
h
an
d
lo
w
th
r
esh
o
ld
s
to
ca
teg
o
r
ize
e
d
g
es a
s
s
tr
o
n
g
o
r
wea
k
.
f)
E
d
g
e
tr
ac
k
i
n
g
b
y
h
y
s
ter
esis
: r
etain
wea
k
ed
g
es e
n
s
u
r
i
n
g
co
n
tin
u
o
u
s
ed
g
e
f
o
r
m
atio
n
.
g)
Ou
tp
u
t:
t
h
e
f
in
al
im
a
g
e
is
g
en
er
ated
,
wh
er
e
e
d
g
es a
r
e
h
ig
h
li
g
h
ted
ag
ai
n
s
t a
b
lack
b
ac
k
g
r
o
u
n
d
.
Af
ter
th
at,
th
e
alg
o
r
ith
m
co
m
p
u
tes
th
e
g
r
ad
ien
t
m
ag
n
itu
d
e
an
d
d
ir
ec
tio
n
at
ea
ch
p
ix
el.
T
h
is
h
elp
s
to
h
ig
h
lig
h
t
ar
ea
s
with
r
a
p
id
i
n
ten
s
ity
ch
an
g
es,
wh
ich
t
y
p
ical
ly
co
r
r
esp
o
n
d
to
ed
g
es.
Fo
llo
win
g
th
e
g
r
ad
ie
n
t
ca
lcu
latio
n
,
n
o
n
-
m
ax
im
u
m
s
u
p
p
r
ess
io
n
is
p
er
f
o
r
m
ed
.
T
h
is
s
tep
id
en
tifie
s
lo
ca
l
m
ax
i
m
a
in
th
e
g
r
ad
ien
t
d
ir
ec
tio
n
a
n
d
s
u
p
p
r
ess
es
n
o
n
-
m
ax
im
u
m
p
i
x
els,
en
s
u
r
in
g
th
at
o
n
ly
th
in
ed
g
es
ar
e
p
r
eser
v
ed
.
T
h
en
,
h
ig
h
a
n
d
lo
w
th
r
esh
o
ld
v
al
u
es
ar
e
d
ef
i
n
ed
to
d
ete
r
m
in
e
wh
ich
ed
g
e
s
ar
e
s
tr
o
n
g
an
d
wh
ich
ar
e
wea
k
.
I
n
th
e
ed
g
e
tr
ac
k
in
g
b
y
h
y
s
ter
esis
s
tep
,
s
t
r
o
n
g
ed
g
es
ab
o
v
e
th
e
h
ig
h
th
r
e
s
h
o
ld
ar
e
m
ar
k
ed
,
an
d
wea
k
ed
g
es
co
n
n
ec
ted
to
s
tr
o
n
g
ed
g
es a
r
e
r
etain
ed
.
Fig
u
r
e
3
.
C
an
n
y
ed
g
e
d
etec
tio
n
tech
n
iq
u
es
T
h
is
h
elp
s
in
co
n
n
ec
tin
g
d
is
co
n
tin
u
o
u
s
ed
g
es
an
d
f
o
r
m
in
g
co
m
p
lete
ed
g
es.
Fin
ally
,
th
e
alg
o
r
ith
m
o
u
tp
u
ts
th
e
f
in
al
ed
g
e
-
d
etec
te
d
im
ag
e,
w
h
er
e
e
d
g
es
ar
e
h
ig
h
lig
h
ted
a
g
ain
s
t
a
b
lack
b
ac
k
g
r
o
u
n
d
.
O
v
er
all,
th
e
C
an
n
y
ed
g
e
d
etec
tio
n
alg
o
r
ith
m
p
r
o
v
id
es
a
r
o
b
u
s
t
m
eth
o
d
f
o
r
ac
cu
r
ately
d
etec
tin
g
e
d
g
es
i
n
im
ag
es,
m
ak
in
g
it
u
s
ef
u
l
in
v
ar
io
u
s
a
p
p
licatio
n
s
s
u
ch
as
o
b
ject
d
etec
tio
n
an
d
im
ag
e
s
eg
m
e
n
tatio
n
.
T
h
e
g
r
a
d
ien
t
m
ag
n
itu
d
e
is
co
m
p
u
ted
u
s
in
g
(
1
)
.
=
√
2
+
2
(
1
)
I
n
th
e
(
1
)
Gx
is
th
e
g
r
ad
ie
n
t
in
x
d
ir
ec
tio
n
,
wh
er
e
Gy
is
th
e
g
r
ad
ien
t
in
y
d
ir
ec
tio
n
.
Als
o
,
g
r
ad
ie
n
t
d
ir
ec
tio
n
is
esti
m
ated
u
s
in
g
th
e
(
2
)
.
=
(
)
(
2
)
3
.
1
.
2
.
P
re
wit
h
edg
e
det
ec
t
io
n
Pre
witt
ed
g
e
d
etec
tio
n
is
a
wi
d
ely
u
s
ed
im
ag
e
p
r
o
ce
s
s
in
g
te
ch
n
iq
u
e
th
at
ap
p
lies
two
3
×
3
k
er
n
els
to
co
m
p
u
te
th
e
g
r
a
d
ien
t
m
ag
n
itu
d
e
in
b
o
th
th
e
x
an
d
y
d
ir
ec
ti
o
n
s
.
T
h
is
h
elp
s
to
h
ig
h
lig
h
t
ed
g
es
in
th
e
i
m
ag
e,
m
ak
in
g
it e
f
f
ec
tiv
e
f
o
r
id
en
tif
y
in
g
s
h
ar
p
ch
an
g
es in
in
ten
s
ity
as illu
s
tr
ated
in
th
e
Fig
u
r
e
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
555
-
1
5
6
8
1560
a)
I
n
p
u
t
im
a
g
e:
co
n
v
er
t
th
e
im
a
g
e
to
g
r
ay
s
ca
le.
b)
Gr
ad
ien
t
ca
lcu
latio
n
: a
p
p
ly
b
o
th
h
o
r
izo
n
tal
an
d
v
er
tical
Pre
witt k
er
n
els to
ca
lcu
late
g
r
ad
ie
n
ts
.
c)
T
h
r
esh
o
ld
in
g
:
s
et
a
th
r
esh
o
ld
t
o
cr
ea
te
a
b
in
a
r
y
im
a
g
e,
en
h
an
cin
g
ed
g
es.
d)
Ou
tp
u
t:
g
en
er
ate
t
h
e
ed
g
e
-
d
et
ec
ted
im
ag
e.
Fig
u
r
e
4
.
Pre
with
ed
g
e
d
etec
ti
o
n
tech
n
i
q
u
es
T
o
p
er
f
o
r
m
Pre
witt
ed
g
e
d
etec
tio
n
,
we
co
n
v
er
t
th
e
im
ag
e
t
o
g
r
ay
s
ca
le,
ap
p
ly
b
o
th
h
o
r
izo
n
tal
an
d
v
er
tical
Pre
witt
k
er
n
els
to
ca
l
cu
late
g
r
ad
ien
ts
,
s
et
a
th
r
esh
o
ld
to
cr
ea
te
a
b
in
ar
y
im
ag
e,
a
n
d
o
u
tp
u
t
th
e
ed
g
e
-
d
etec
ted
im
ag
e.
T
h
e
h
o
r
izo
n
ta
l g
r
ad
ien
t (
G
x
)
is
g
iv
en
in
th
e
(
3
)
:
=
(
−
1
0
1
−
1
0
1
−
1
0
1
)
(
3
)
t
h
e
v
er
tical
g
r
a
d
ien
t (
Gy
)
is
s
h
o
wn
in
t
h
e
(
4
)
.
=
(
−
1
−
1
1
0
0
0
1
1
1
)
(
4
)
T
h
e
f
lo
wch
a
r
t
s
h
o
wn
in
Fig
u
r
e
4
th
e
s
tep
s
in
v
o
l
v
ed
i
n
Pre
witt
ed
g
e
d
etec
tio
n
,
f
r
o
m
im
a
g
e
in
p
u
t
to
o
b
tain
in
g
th
e
f
in
al
ed
g
e
-
d
ete
cted
r
esu
lt.
Pre
witt
ed
g
e
d
et
ec
tio
n
is
a
s
im
p
ler
tech
n
iq
u
e
th
an
C
an
n
y
ed
g
e
d
etec
tio
n
.
Pre
witt
in
v
o
lv
es
co
n
v
o
lv
in
g
th
e
im
ag
e
with
two
f
ix
ed
k
er
n
els.
C
an
n
y
u
s
es
s
eq
u
en
tial
o
p
er
atio
n
s
lik
e
Gau
s
s
ian
b
lu
r
r
in
g
,
g
r
ad
ien
t
co
m
p
u
tatio
n
,
n
o
n
-
m
a
x
im
u
m
s
u
p
p
r
ess
io
n
,
an
d
h
y
s
ter
esis
th
r
esh
o
ld
in
g
.
C
an
n
y
p
r
o
v
id
es m
o
r
e
ac
cu
r
ate
ed
g
e
d
etec
tio
n
with
r
ed
u
ce
d
n
o
is
e
an
d
b
ette
r
b
o
u
n
d
ar
y
id
en
t
if
icatio
n
.
3
.
1
.
3
.
So
bel
edg
e
det
ec
t
io
n
So
b
el
ed
g
e
d
etec
tio
n
is
a
te
ch
n
iq
u
e
u
s
ed
to
d
etec
t
e
d
g
e
s
in
im
ag
es
b
y
co
m
p
u
tin
g
t
h
e
g
r
ad
ien
t
ap
p
r
o
x
im
atio
n
.
I
t
in
v
o
lv
es
co
n
v
o
lv
in
g
th
e
im
ag
e
with
a
p
ai
r
o
f
3
×
3
k
er
n
els
to
ca
lcu
late
th
e
g
r
a
d
ien
ts
in
t
h
e
h
o
r
izo
n
tal
an
d
v
e
r
tical
d
ir
ec
tio
n
s
,
wh
ich
ar
e
th
en
co
m
b
in
e
d
to
d
eter
m
in
e
th
e
m
ag
n
itu
d
e
an
d
o
r
ien
tatio
n
o
f
ed
g
es.
T
h
is
m
eth
o
d
is
s
im
p
l
e,
ef
f
ec
tiv
e
,
an
d
wid
el
y
u
s
ed
in
o
b
ject
d
etec
tio
n
an
d
i
m
ag
e
s
eg
m
e
n
tatio
n
.
T
h
e
f
lo
w
ch
a
r
t o
f
So
b
el
ed
g
e
d
etec
tio
n
is
p
r
esen
ted
in
Fig
u
r
e
5
.
a)
I
n
p
u
t
im
a
g
e:
co
n
v
er
t
th
e
im
a
g
e
to
g
r
ay
s
ca
le.
b)
Ho
r
izo
n
tal
g
r
a
d
ien
t: a
p
p
l
y
a
h
o
r
izo
n
tal
So
b
el
k
er
n
el
to
d
etec
t h
o
r
izo
n
tal
g
r
ad
ien
ts
.
c)
Ver
tical
g
r
ad
ien
t: a
p
p
l
y
a
v
e
r
tical
So
b
el
k
er
n
el
to
d
etec
t v
er
ti
ca
l g
r
ad
ien
ts
.
d)
Gr
ad
ien
t
m
ag
n
itu
d
e:
co
m
p
u
te
th
e
g
r
ad
ien
t
m
ag
n
itu
d
es.
e)
T
h
r
esh
o
ld
in
g
:
d
ef
in
e
a
th
r
esh
o
ld
to
en
h
a
n
ce
ed
g
es.
f)
Ou
tp
u
t:
p
r
o
d
u
ce
th
e
ed
g
e
-
d
ete
cted
im
ag
e.
Fig
u
r
e
5
.
So
b
el
ed
g
e
d
etec
tio
n
tech
n
iq
u
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
C
o
mp
a
r
a
tive
a
n
a
lysi
s
o
f e
d
g
e
d
etec
tio
n
ima
g
e
p
r
o
ce
s
s
in
g
… (
N
ir
a
n
ja
n
a
C
h
a
n
d
r
a
s
ek
a
r
a
B
h
a
r
a
th
y
)
1561
B
eg
in
b
y
in
p
u
ttin
g
an
im
ag
e
an
d
co
n
v
er
tin
g
it
to
g
r
ay
s
ca
le
to
f
ac
ilit
ate
p
r
o
ce
s
s
in
g
.
Ap
p
ly
a
h
o
r
izo
n
tal
So
b
el
k
er
n
el
th
r
o
u
g
h
co
n
v
o
lu
tio
n
to
d
etec
t
h
o
r
izo
n
tal
g
r
ad
ien
ts
,
f
o
llo
wed
b
y
a
v
er
tical
So
b
el
k
er
n
el
to
id
e
n
tify
v
e
r
tical
g
r
ad
ien
ts
.
C
o
m
p
u
te
th
e
g
r
a
d
ien
t m
ag
n
itu
d
es f
o
r
ea
ch
p
i
x
el
an
d
e
s
tab
lis
h
a
th
r
esh
o
ld
v
alu
e.
Set
p
ix
els
with
g
r
ad
ien
t
m
ag
n
itu
d
es
ab
o
v
e
th
e
th
r
esh
o
ld
to
wh
ite
an
d
th
o
s
e
b
elo
w
to
b
lack
.
C
o
n
clu
d
e
b
y
o
u
t
p
u
ttin
g
th
e
e
d
g
e
-
d
etec
ted
im
ag
e.
T
h
is
p
r
o
ce
s
s
en
h
a
n
ce
s
ed
g
es
in
th
e
im
ag
e,
m
a
k
in
g
th
em
s
t
an
d
o
u
t
ag
ain
s
t th
e
b
ac
k
g
r
o
u
n
d
.
T
h
e
h
o
r
izo
n
tal
an
d
v
er
tical
s
o
b
el
k
e
r
n
el
is
s
h
o
wn
in
th
e
(
5
)
a
n
d
(
6
)
r
esp
ec
tiv
ely
.
=
(
−
1
0
1
−
2
0
2
−
1
0
1
)
(
6
)
=
(
−
1
−
2
−
1
0
0
0
1
2
1
)
(
7
)
So
b
el
ed
g
e
d
etec
tio
n
is
a
s
im
p
ler
tech
n
iq
u
e
th
at
ca
lcu
lates
g
r
ad
ien
ts
u
s
in
g
f
ix
ed
So
b
el
k
e
r
n
els
in
th
e
h
o
r
izo
n
tal
an
d
v
er
tical
d
ir
ec
ti
o
n
s
.
I
t
’
s
co
m
p
u
tatio
n
ally
e
f
f
i
cien
t
b
u
t
ca
n
b
e
s
en
s
itiv
e
to
n
o
is
e,
r
esu
ltin
g
in
th
ick
er
ed
g
es.
C
o
n
v
er
s
ely
,
C
an
n
y
ed
g
e
d
etec
tio
n
is
m
o
r
e
s
o
p
h
is
ticated
,
em
p
lo
y
in
g
m
u
l
tip
le
s
tep
s
s
u
ch
as
Gau
s
s
ian
b
lu
r
r
in
g
,
g
r
ad
ien
t
c
o
m
p
u
tatio
n
,
n
o
n
-
m
a
x
im
u
m
s
u
p
p
r
ess
io
n
,
a
n
d
h
y
s
ter
esis
th
r
esh
o
ld
in
g
.
C
an
n
y
ed
g
e
d
etec
tio
n
p
r
o
d
u
ce
s
th
i
n
n
er
an
d
m
o
r
e
ac
cu
r
ate
e
d
g
es,
m
ak
in
g
it
s
u
itab
le
f
o
r
ap
p
lica
tio
n
s
wh
er
e
p
r
ec
is
e
ed
g
e
lo
ca
lizatio
n
an
d
n
o
is
e
to
ler
an
ce
ar
e
cr
u
cial,
alb
eit
with
a
h
ig
h
er
c
o
m
p
u
tatio
n
al
co
s
t.
I
n
co
n
t
r
ast,
So
b
el
ed
g
e
d
etec
tio
n
is
f
aster
an
d
m
o
r
e
s
tr
aig
h
tf
o
r
war
d
,
m
a
k
in
g
it su
itab
le
f
o
r
s
im
p
ler
ed
g
e
d
etec
tio
n
task
s
.
3
.
1
.
4
.
Ro
bert
edg
e
det
ec
t
io
n
R
o
b
er
t
ed
g
e
d
etec
tio
n
is
an
i
m
ag
e
p
r
o
ce
s
s
in
g
tech
n
i
q
u
e
t
h
at
id
en
tifie
s
ed
g
es
b
y
a
p
p
ly
in
g
a
s
p
ec
if
ic
k
er
n
el
to
th
e
im
a
g
e.
W
h
ile
s
im
ilar
to
So
b
el
ed
g
e
d
etec
tio
n
,
it
u
tili
ze
s
a
d
is
tin
ct
k
er
n
el.
T
h
e
R
o
b
er
t
ed
g
e
d
etec
tio
n
m
eth
o
d
em
p
l
o
y
s
two
2
×
2
m
atr
ices,
wh
ich
ar
e
co
n
v
o
lv
ed
with
th
e
im
ag
e
to
co
m
p
u
te
h
o
r
izo
n
tal
an
d
v
er
tical
g
r
ad
ien
ts
.
T
h
ese
g
r
ad
i
en
ts
ar
e
th
en
co
m
b
in
e
d
to
d
et
ec
t
ed
g
es
in
th
e
im
ag
e.
Du
e
to
its
s
im
p
licity
an
d
co
m
p
u
tatio
n
al
ef
f
icien
cy
,
R
o
b
er
t
ed
g
e
d
etec
tio
n
is
well
-
s
u
ited
f
o
r
r
ea
l
-
tim
e
ed
g
e
d
etec
tio
n
ap
p
licatio
n
s
.
T
h
e
f
lo
w
ch
ar
t
o
f
R
o
b
er
t
ed
g
e
d
etec
tio
n
tech
n
iq
u
e
is
s
h
o
wn
in
Fi
g
u
r
e
6
.
T
h
e
ex
p
r
ess
io
n
f
o
r
h
o
r
izo
n
tal
R
o
b
er
t
k
er
n
el
an
d
v
er
tical
R
o
b
er
t
k
er
n
el
is
s
h
o
wn
in
th
e
(
8
)
an
d
(
9
)
r
esp
ec
tiv
ely
.
=
(
1
0
0
−
1
)
(
8
)
=
(
0
1
−
1
0
)
(
9
)
Fig
u
r
e
6
.
R
o
b
e
r
t e
d
g
e
d
etec
tio
n
tech
n
iq
u
e
E
d
g
e
d
etec
tio
n
tech
n
i
q
u
es
s
u
c
h
as
R
o
b
er
t,
So
b
el,
an
d
Pre
witt
ar
e
u
s
ed
to
d
etec
t
ed
g
es
in
an
im
ag
e.
R
o
b
er
t
an
d
o
th
er
b
asic
tech
n
i
q
u
es
wo
r
k
b
y
co
n
v
o
l
v
in
g
th
e
i
m
ag
e
with
s
p
ec
if
ic
k
er
n
els
to
d
etec
t
ed
g
es,
wh
ich
m
ak
es
th
em
co
m
p
u
tatio
n
ally
s
im
p
ler
an
d
f
aster
.
T
h
ey
m
ay
p
r
o
d
u
ce
th
ic
k
er
ed
g
es
an
d
b
e
s
en
s
itiv
e
to
n
o
is
e.
Alter
n
ativ
ely
,
t
h
e
C
an
n
y
e
d
g
e
d
etec
tio
n
alg
o
r
ith
m
r
e
p
r
esen
ts
an
ad
v
an
ce
d
ap
p
r
o
ac
h
en
co
m
p
ass
in
g
s
ev
er
al
p
h
ases
:
Gau
s
s
ian
s
m
o
o
th
in
g
,
g
r
ad
ien
t
co
m
p
u
tatio
n
,
s
u
p
p
r
es
s
io
n
o
f
n
o
n
-
m
a
x
im
a,
an
d
t
h
r
e
s
h
o
ld
d
eter
m
in
atio
n
v
ia
h
y
s
ter
esis
.
I
t
p
r
o
v
i
d
es
s
u
p
er
io
r
ed
g
e
d
etec
tio
n
p
er
f
o
r
m
a
n
ce
b
y
p
r
o
d
u
cin
g
t
h
in
n
er
an
d
m
o
r
e
p
r
ec
is
e
ed
g
es,
m
ak
in
g
it su
itab
le
f
o
r
ap
p
licati
o
n
s
wh
er
e
ac
cu
r
ate
e
d
g
e
lo
ca
li
za
tio
n
an
d
n
o
is
e
r
ed
u
ctio
n
ar
e
cr
u
cial.
Ho
wev
er
,
it
r
eq
u
ir
es
m
o
r
e
c
o
m
p
u
tatio
n
a
l
r
eso
u
r
ce
s
.
I
n
s
u
m
m
ar
y
,
w
h
ile
b
asic
tech
n
iq
u
es
lik
e
R
o
b
er
t
,
So
b
el,
an
d
Pre
witt
a
r
e
f
a
s
t
e
r
a
n
d
s
i
m
p
l
e
r
,
C
a
n
n
y
p
r
o
v
i
d
e
s
b
e
t
t
e
r
p
e
r
f
o
r
m
a
n
c
e
a
t
t
h
e
c
o
s
t
o
f
i
n
c
r
e
a
s
e
d
c
o
m
p
l
e
x
i
t
y
a
n
d
c
o
m
p
u
t
a
t
i
o
n
a
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
555
-
1
5
6
8
1562
r
eq
u
ir
em
e
n
ts
.
T
h
is
p
r
o
ce
s
s
en
h
an
ce
s
ed
g
es
in
th
e
im
ag
e,
m
ak
in
g
t
h
em
s
tan
d
o
u
t
ag
ain
s
t
th
e
b
ac
k
g
r
o
u
n
d
.
R
o
b
er
t e
d
g
e
d
etec
tio
n
u
s
es 2
×
2
k
er
n
els to
ca
lc
u
late
g
r
ad
ie
n
ts
:
a)
I
n
p
u
t
im
a
g
e:
co
n
v
er
t
th
e
im
a
g
e
to
g
r
ay
s
ca
le.
b)
Gr
ad
ien
t
ca
lcu
latio
n
: a
p
p
ly
h
o
r
izo
n
tal
an
d
v
er
tical
R
o
b
er
t k
e
r
n
els to
ca
lcu
late
g
r
a
d
ien
ts
.
c)
T
h
r
esh
o
ld
in
g
:
c
o
m
b
i
n
e
g
r
a
d
ie
n
ts
to
id
en
tify
e
d
g
es.
d)
Ou
tp
u
t:
g
en
er
ate
t
h
e
ed
g
e
-
d
et
ec
ted
im
ag
e.
e)
T
h
e
f
lo
wch
ar
t
f
o
r
r
o
b
e
r
t e
d
g
e
d
etec
tio
n
is
s
h
o
wn
in
Fig
u
r
e
6
3
.
2
.
I
m
ple
m
ent
a
t
io
n o
f
I
m
a
g
e
pro
ce
s
s
ing
t
ec
hn
iq
ue
in o
penCV
env
iro
nm
ent
I
n
o
r
d
e
r
to
co
m
p
ar
e
p
o
p
u
lar
ed
g
e
d
etec
tio
n
tech
in
iq
u
es
an
d
im
ag
e
s
eg
m
en
tatio
n
tech
n
i
q
u
es,
ea
ch
alg
o
r
ith
m
s
ar
e
d
ev
elo
p
e
d
in
o
p
en
C
V
p
latf
o
r
m
u
s
in
g
p
y
th
o
n
p
r
o
g
r
am
m
in
g
.
T
h
e
s
tep
s
o
f
im
p
lem
en
tatio
n
ar
e
d
escr
ib
ed
h
e
r
e.
E
d
g
es
th
at
a
r
e
co
n
n
ec
ted
to
s
tr
o
n
g
ed
g
e
s
wi
th
a
m
ag
n
itu
d
e
a
b
o
v
e
t
h
r
esh
o
ld
1
ar
e
also
co
n
s
id
er
ed
v
alid
e
d
g
es.
Fin
ally
,
d
is
p
lay
b
o
t
h
th
e
o
r
ig
i
n
al
an
d
th
e
ed
g
e
-
d
etec
ted
im
ag
e
u
s
i
n
g
cv
2
.
im
s
h
o
w
an
d
t
h
en
,
p
a
u
s
e
f
o
r
u
s
er
in
p
u
t
t
o
te
r
m
in
ate
th
e
win
d
o
ws u
s
in
g
cv
2
.
waitKe
y
an
d
c
v
2
.
d
estro
y
AllW
in
d
o
ws.
T
h
e
alg
o
r
ith
m
f
o
r
im
p
lem
e
n
tin
g
th
e
ed
g
e
d
etec
tio
n
tech
n
iq
u
e
i
n
o
p
en
C
V
p
latf
o
r
m
is
b
r
ief
ly
illu
s
tr
ated
h
er
e.
C
o
n
s
id
er
th
e
s
tep
s
f
o
r
th
e
im
p
lem
en
tatio
n
:
a)
I
m
p
o
r
t O
p
e
n
C
V:
i
m
p
o
r
t th
e
O
p
en
C
V
lib
r
ar
y
(
c
v
2
)
.
b)
L
o
ad
im
ag
e:
ac
q
u
ir
e
th
e
in
p
u
t
im
ag
e
v
ia
cv
2
.
im
r
ea
d
.
c)
Gr
ay
s
ca
le
co
n
v
er
s
io
n
:
tr
a
n
s
f
o
r
m
th
e
im
ag
e
to
g
r
ay
s
ca
le
u
tili
zin
g
cv
2
.
cv
tC
o
lo
r
.
d)
No
is
e
r
ed
u
ctio
n
:
im
p
lem
e
n
t
G
au
s
s
ian
b
lu
r
th
r
o
u
g
h
cv
2
.
Ga
u
s
s
ian
B
lu
r
to
m
itig
ate
n
o
is
e
.
e)
E
d
g
e
d
etec
tio
n
:
i
m
p
lem
en
t
th
e
ch
o
s
en
ed
g
e
d
etec
tio
n
m
eth
o
d
u
s
in
g
th
e
r
esp
ec
tiv
e
Op
en
C
V
f
u
n
ctio
n
s
.
f)
T
h
r
esh
o
ld
in
g
:
f
o
r
C
an
n
y
,
s
et
T
h
r
esh
o
ld
1
(
lo
wer
th
r
esh
o
ld
)
an
d
T
h
r
esh
o
ld
2
(
u
p
p
er
th
r
esh
o
ld
)
.
g)
Dis
p
lay
r
esu
lts
:
d
i
s
p
lay
b
o
th
th
e
o
r
ig
in
al
an
d
ed
g
e
-
d
etec
ted
im
ag
es
u
s
in
g
cv
2
.
im
s
h
o
w.
Use
cv
2
.
waitKe
y
an
d
cv
2
.
d
estro
y
AllW
in
d
o
ws to
m
an
a
g
e
win
d
o
w
d
is
p
lay
.
3
.
3
.
E
x
perim
ent
a
l
s
et
up
T
h
e
ex
p
er
im
e
n
tal
s
etu
p
f
o
r
o
u
r
r
esear
ch
o
n
e
d
g
e
d
etec
tio
n
tech
n
iq
u
es
in
tr
af
f
ic
s
ig
n
al
m
an
ag
em
en
t
in
v
o
lv
ed
m
u
ltip
le
s
tag
es,
en
co
m
p
ass
in
g
v
ar
io
u
s
m
eth
o
d
o
lo
g
ies,
in
s
tr
u
m
en
ts
,
m
ater
ials
,
p
r
o
ce
d
u
r
es,
an
d
m
ea
s
u
r
em
en
ts
.
T
h
e
p
r
im
ar
y
in
s
tr
u
m
en
t
u
s
ed
f
o
r
ca
p
tu
r
in
g
tr
af
f
ic
im
ag
es
was
a
h
i
g
h
-
r
eso
lu
tio
n
ca
m
e
r
a,
s
tr
ateg
ically
p
o
s
itio
n
ed
at
m
u
ltip
le
in
ter
s
ec
tio
n
s
to
ca
p
tu
r
e
clea
r
im
ag
es
o
f
v
eh
icles
an
d
p
ed
estrian
s
.
T
h
ese
im
ag
es
wer
e
th
e
n
t
r
an
s
f
o
r
m
e
d
u
s
in
g
th
e
O
p
en
C
V
in
a
Py
th
o
n
p
r
o
g
r
a
m
m
in
g
en
v
ir
o
n
m
en
t.
T
h
e
ed
g
e
d
etec
tio
n
tech
n
iq
u
es
ev
alu
ated
in
cl
u
d
e
d
C
an
n
y
,
Pre
witt,
So
b
el,
an
d
R
o
b
er
t
m
eth
o
d
s
.
E
ac
h
ed
g
e
d
etec
tio
n
tech
n
iq
u
e
was
im
p
lem
en
ted
b
y
p
r
o
c
u
r
in
g
th
e
Op
en
C
V
lib
r
ar
y
an
d
s
t
o
ck
in
g
th
e
tr
af
f
ic
im
ag
es
em
p
lo
y
in
g
c
v
2
.
im
r
ea
d
.
I
m
ag
es
wer
e
o
p
tio
n
ally
c
o
n
v
er
ted
to
g
r
ay
s
ca
le
u
s
in
g
cv
2
.
cv
tC
o
lo
r
to
s
im
p
lify
p
r
o
ce
s
s
in
g
.
Gau
s
s
ian
b
lu
r
(
cv
2
.
Gau
s
s
ian
B
lu
r
)
was
ap
p
li
ed
to
r
e
d
u
ce
n
o
is
e,
with
th
e
k
er
n
el
s
ize
ad
ju
s
ted
b
ased
o
n
t
h
e
im
ag
e
p
r
o
p
er
ties
.
T
h
e
co
r
e
ed
g
e
d
etec
tio
n
was
p
er
f
o
r
m
ed
u
s
in
g
th
e
r
esp
ec
tiv
e
Op
en
C
V
f
u
n
ctio
n
s
f
o
r
ea
ch
tech
n
iq
u
e,
s
u
ch
a
s
cv
2
.
C
an
n
y
f
o
r
C
an
n
y
ed
g
e
d
etec
tio
n
,
wh
er
e
two
th
r
esh
o
l
d
v
alu
es
wer
e
s
p
ec
if
ied
to
d
i
f
f
er
en
tiate
b
etwe
en
s
tr
o
n
g
an
d
wea
k
ed
g
es.
T
h
e
o
u
tp
u
t
im
ag
es
wer
e
d
is
p
lay
ed
u
s
in
g
cv
2
.
im
s
h
o
w,
a
n
d
u
s
er
i
n
p
u
ts
wer
e
h
an
d
led
with
cv
2
.
waitKe
y
an
d
cv
2
.
d
estro
y
AllW
in
d
o
ws
f
o
r
clo
s
in
g
win
d
o
ws.
E
ac
h
tech
n
iq
u
e
’
s
p
er
f
o
r
m
an
ce
was
ev
alu
ated
b
ased
o
n
s
ev
er
al
cr
iter
ia,
in
clu
d
in
g
o
b
ject
s
iz
e
d
etec
tio
n
,
en
v
ir
o
n
m
en
tal
l
ig
h
tin
g
co
n
d
itio
n
s
,
m
u
ltip
le
d
etec
tio
n
c
ap
ab
ilit
ies,
an
d
p
r
o
ce
s
s
in
g
tim
e.
T
h
e
C
am
er
a
s
p
ec
if
icatio
n
s
is
Hik
v
is
io
n
DS
-
2
C
D2
0
8
5
FW
D
-
I
m
o
d
el,
8
M
P
(
3
8
4
0
×
2
1
6
0
)
R
eso
lu
tio
n
an
d
3
0
f
r
am
es
p
er
s
ec
o
n
d
(
FPS
)
is
th
e
f
r
am
e
r
ate.
T
h
e
C
am
er
as
wer
e
s
tr
ateg
ically
p
lace
d
at
v
ar
io
u
s
in
ter
s
ec
ti
o
n
s
to
co
v
er
d
if
f
er
en
t
la
n
es
o
f
tr
af
f
ic,
e
n
s
u
r
in
g
a
co
m
p
r
eh
e
n
s
iv
e
v
iew
o
f
v
e
h
icle
an
d
p
ed
estrian
m
o
v
em
en
ts
.
T
h
e
o
r
ien
tatio
n
was
ad
ju
s
ted
to
ca
p
tu
r
e
th
e
m
ax
im
u
m
f
ield
o
f
v
iew
with
o
u
t
o
b
s
tr
u
ctio
n
s
,
ty
p
ically
m
o
u
n
ted
o
n
p
o
les
o
r
b
u
ild
in
g
co
r
n
er
s
at
a
h
eig
h
t
o
f
15
-
2
0
f
ee
t.
I
m
ag
es
wer
e
ca
p
t
u
r
ed
at
a
f
r
eq
u
en
cy
o
f
1
f
p
s
to
b
alan
ce
d
ata
g
r
a
n
u
lar
ity
a
n
d
p
r
o
ce
s
s
in
g
lo
ad
u
n
d
er
v
ar
io
u
s
lig
h
tin
g
co
n
d
itio
n
s
,
in
clu
d
in
g
b
r
i
g
h
t
d
ay
lig
h
t
,
d
u
s
k
,
an
d
n
ig
h
t
-
tim
e
with
s
tr
ee
t
lig
h
tin
g
.
T
h
is
en
s
u
r
ed
th
at
th
e
ed
g
e
d
etec
ti
o
n
tech
n
iq
u
es
wer
e
test
ed
ac
r
o
s
s
d
if
f
er
en
t
en
v
ir
o
n
m
en
tal
lig
h
tin
g
s
ce
n
ar
io
s
.
I
m
ag
es
wer
e
r
esized
to
a
s
tan
d
ar
d
r
eso
lu
tio
n
o
f
1
9
2
0
×
1
0
8
0
p
i
x
els
to
m
ain
tai
n
co
n
s
is
t
en
cy
in
p
r
o
ce
s
s
in
g
.
Har
d
war
e
in
clu
d
es
I
n
tel
C
o
r
e
i9
-
9
9
0
0
K
p
r
o
ce
s
s
o
r
is
u
s
ed
,
3
2
GB
DDR4
R
AM
,
NVI
DI
A
GeFo
r
ce
R
T
X
2
0
8
0
T
i
GPU.
So
f
twar
e
in
clu
d
es
Op
e
n
C
V
v
er
s
io
n
4
.
5
.
1
f
o
r
im
p
lem
en
tin
g
an
d
test
in
g
th
e
ed
g
e
d
etec
tio
n
alg
o
r
ith
m
s
.
Ot
h
er
s
o
f
twar
e
o
r
to
o
ls
u
s
ed
ar
e
Py
t
h
o
n
3
.
8
f
o
r
s
cr
ip
tin
g
an
d
alg
o
r
ith
m
d
ev
elo
p
m
en
t.
T
h
e
i
m
a
g
e
s
w
e
r
e
s
el
f
-
c
o
l
l
e
ct
e
d
u
s
i
n
g
t
h
e
a
f
o
r
e
m
e
n
t
i
o
n
e
d
c
a
m
e
r
a
s
e
t
u
p
a
t
m
u
l
t
i
p
l
e
i
n
t
e
r
s
ec
t
i
o
n
s
i
n
a
m
e
t
r
o
p
o
l
i
t
a
n
a
r
e
a
o
v
e
r
a
p
e
r
i
o
d
o
f
o
n
e
m
o
n
t
h
.
A
t
o
t
a
l
o
f
1
0
,
0
0
0
i
m
a
g
e
s
w
e
r
e
c
o
l
le
c
t
e
d
a
n
d
u
s
e
d
f
o
r
t
h
e
a
n
a
l
y
s
is
.
T
h
e
d
ataset
in
cl
u
d
ed
im
ag
es
with
v
ar
y
i
n
g
t
r
af
f
ic
d
en
s
ities
,
r
an
g
in
g
f
r
o
m
lo
w
(
f
ew
v
e
h
icles)
to
h
ig
h
(
h
ea
v
y
c
o
n
g
esti
o
n
)
.
I
m
a
g
es
wer
e
s
to
r
ed
in
PNG
f
o
r
m
at
to
p
r
eser
v
e
q
u
ality
a
n
d
s
u
p
p
o
r
t
lo
s
s
les
s
co
m
p
r
ess
io
n
.
T
h
e
s
tan
d
ar
d
ized
im
ag
e
s
ize
was
1
9
2
0
×
1
0
8
0
p
ix
els,
en
s
u
r
in
g
u
n
if
o
r
m
ity
f
o
r
p
r
o
ce
s
s
in
g
.
T
h
e
d
ata
an
aly
s
is
f
o
cu
s
ed
o
n
ass
ess
in
g
th
e
p
er
f
o
r
m
an
ce
o
f
ea
ch
e
d
g
e
d
etec
tio
n
tech
n
iq
u
e
ag
ai
n
s
t
v
ar
io
u
s
p
ar
am
eter
s
,
an
d
d
ec
is
io
n
s
f
o
r
i
n
clu
d
in
g
o
r
ex
clu
d
in
g
ce
r
tain
d
ata
wer
e
b
a
s
ed
o
n
p
r
ed
ef
i
n
ed
c
r
iter
ia.
T
r
af
f
ic
im
ag
es
wer
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
C
o
mp
a
r
a
tive
a
n
a
lysi
s
o
f e
d
g
e
d
etec
tio
n
ima
g
e
p
r
o
ce
s
s
in
g
… (
N
ir
a
n
ja
n
a
C
h
a
n
d
r
a
s
ek
a
r
a
B
h
a
r
a
th
y
)
1563
co
llected
u
n
d
er
d
if
f
e
r
en
t
e
n
v
ir
o
n
m
e
n
tal
co
n
d
itio
n
s
a
n
d
with
v
ar
y
i
n
g
t
r
af
f
ic
d
en
s
ities
.
T
h
e
p
er
f
o
r
m
an
ce
m
etr
ics
in
clu
d
ed
e
d
g
e
s
h
ar
p
n
ess
,
ab
r
u
p
tn
ess
,
s
en
s
itiv
ity
f
ac
to
r
s
(
SF
)
,
an
d
p
r
o
ce
s
s
in
g
ti
m
e.
A
co
m
p
ar
is
o
n
m
atr
ix
was
f
o
r
m
u
lated
to
s
y
s
tem
atica
lly
ev
alu
ate
t
h
ese
m
et
r
ics,
en
s
u
r
in
g
a
c
o
m
p
r
eh
en
s
iv
e
an
aly
s
is
.
C
er
tain
d
ata
p
o
in
ts
wer
e
ex
clu
d
e
d
if
t
h
ey
d
id
n
o
t
m
ee
t
th
e
q
u
ality
s
tan
d
ar
d
s
r
e
q
u
ir
e
d
f
o
r
ac
cu
r
at
e
an
aly
s
is
,
s
u
ch
as
im
ag
es
with
ex
ce
s
s
iv
e
n
o
is
e
o
r
th
o
s
e
tak
en
u
n
d
er
e
x
tr
e
m
e
lig
h
tin
g
co
n
d
itio
n
s
th
at
r
en
d
er
e
d
th
e
ed
g
e
d
etec
tio
n
in
ef
f
ec
tiv
e
.
T
h
e
d
ata
was
p
r
o
ce
s
s
ed
an
d
an
aly
ze
d
u
s
in
g
Py
th
o
n
s
cr
ip
ts
,
en
s
u
r
in
g
co
n
s
is
ten
t
an
d
r
ep
ea
tab
le
r
e
s
u
lts
.
T
h
e
in
clu
s
io
n
cr
iter
ia
f
o
c
u
s
ed
o
n
im
a
g
es
th
at
r
ep
r
esen
te
d
ty
p
ical
tr
af
f
ic
co
n
d
itio
n
s
,
en
ab
lin
g
a
r
o
b
u
s
t
co
m
p
ar
is
o
n
o
f
ed
g
e
d
etec
tio
n
tech
n
iq
u
es.
T
h
e
f
in
d
in
g
s
wer
e
tab
u
lated
an
d
an
aly
ze
d
to
d
eter
m
in
e
th
e
m
o
s
t
ef
f
ec
tiv
e
m
eth
o
d
f
o
r
tr
af
f
ic
s
ig
n
al
m
a
n
a
g
em
en
t,
with
th
e
C
an
n
y
ed
g
e
d
etec
tio
n
tech
n
iq
u
e
s
h
o
win
g
s
u
p
er
io
r
p
er
f
o
r
m
a
n
c
e
in
m
o
s
t
ev
alu
ated
p
ar
am
et
er
s
.
T
h
is
s
tep
h
elp
s
in
m
in
i
m
izin
g
f
alse
ed
g
e
d
etec
tio
n
.
T
h
e
k
er
n
el
s
ize
f
o
r
Gau
s
s
ian
b
lu
r
was
em
p
ir
ically
s
et
to
5
×
5
b
ased
o
n
p
r
elim
in
ar
y
ex
p
er
im
en
ts
.
Fo
r
n
o
r
m
aliza
tio
n
all
im
ag
e
s
ar
e
co
n
v
er
ted
t
o
g
r
ay
s
ca
l
e
u
s
in
g
cv
2
.
c
v
tC
o
lo
r
.
T
h
is
s
tep
s
im
p
lifie
d
th
e
p
r
o
ce
s
s
in
g
b
y
r
e
d
u
cin
g
th
e
c
o
m
p
lex
ity
f
r
o
m
th
r
ee
co
lo
r
c
h
an
n
els
to
o
n
e
an
d
R
esized
im
ag
es
to
a
s
tan
d
ar
d
r
eso
lu
tio
n
o
f
1
9
2
0
×
1
0
8
0
p
ix
e
ls
to
en
s
u
r
e
u
n
if
o
r
m
ity
in
s
u
b
s
eq
u
en
t
p
r
o
ce
s
s
in
g
.
E
v
alu
ated
th
e
co
r
r
ec
tn
ess
o
f
ed
g
e
d
etec
tio
n
b
y
ev
alu
atin
g
o
b
s
er
v
e
d
e
d
g
es
with
m
an
u
a
lly
an
n
o
tated
g
r
o
u
n
d
tr
u
t
h
e
d
g
es
f
o
r
ac
c
u
r
ac
y
.
T
h
e
p
r
o
p
o
r
tio
n
o
f
co
r
r
ec
tly
id
en
tifie
d
ed
g
es
to
th
e
to
tal
n
u
m
b
er
o
f
ed
g
es
id
en
tifie
d
.
T
h
e
p
r
o
p
o
r
tio
n
o
f
co
r
r
ec
tly
id
e
n
tifie
d
ed
g
es to
th
e
to
tal
n
u
m
b
e
r
o
f
ac
tu
al
ed
g
es
in
th
e
g
r
o
u
n
d
tr
u
th
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
Fo
r
th
e
s
ak
e
o
f
an
aly
s
e
th
e
ev
alu
atio
n
o
f
ed
g
e
d
etec
ti
o
n
tech
n
i
q
u
e
an
d
im
ag
e
s
eg
m
en
tatio
n
tech
n
iq
u
e,
tr
af
f
ic
im
ag
es
wer
e
co
llected
.
A
co
m
p
ar
is
o
n
m
atr
ix
is
f
o
r
m
u
lated
w
h
ich
in
cl
u
d
es
o
r
ien
tatio
n
,
s
ize,
en
v
ir
o
n
m
en
tal
lig
h
tin
g
,
m
u
lti
p
le
d
etec
tio
n
,
tr
a
f
f
ic
d
en
s
ity
,
s
h
ar
p
n
ess
r
atio
(
SR
)
,
ab
r
u
p
tn
ess
,
an
d
SF
a
r
e
co
n
s
id
er
ed
f
o
r
th
e
co
m
p
ar
ativ
e
an
aly
s
is
.
T
ab
le
1
s
h
o
ws
th
e
co
m
p
ar
ativ
e
an
aly
s
is
o
f
ty
p
e
s
o
f
ed
g
e
d
etec
tio
n
tech
n
iq
u
e
em
p
lo
y
e
d
f
o
r
th
e
T
MS.
T
h
e
f
ir
s
t
p
ar
am
eter
f
o
r
th
e
co
m
p
ar
i
o
n
o
f
ed
g
e
d
etec
tio
n
tech
n
iq
u
e
is
b
ased
o
n
th
e
s
ize
o
f
th
e
o
b
ject
d
etec
tio
n
.
I
t
ca
n
b
e
o
b
s
er
v
ed
th
at
s
m
all,
m
ed
iu
m
,
an
d
lar
g
e
s
ize
o
b
ject
d
etec
tio
n
is
p
o
s
s
ib
le
u
s
in
g
all
th
e
ed
g
e
d
etec
tio
n
tech
n
iq
u
e.
All
th
e
o
b
j
ec
ts
co
n
s
id
er
ed
f
o
r
th
e
an
aly
s
is
ar
e
u
n
f
am
ilar
iar
,
ie
n
o
p
r
e
-
d
ef
i
n
ed
f
ea
tu
r
es
ar
e
av
ialb
le
r
eg
ar
d
in
g
th
e
n
atu
r
e
o
f
th
e
o
b
ject.
T
h
er
e
ar
e
t
h
r
ee
s
u
b
d
iv
is
io
n
o
f
en
v
ir
o
n
m
en
tal
lig
h
tin
g
s
u
ch
as
b
r
ig
h
t,e
n
o
u
g
h
an
d
d
u
ll
Fig
u
r
e
7
.
T
h
e
C
an
n
y
ed
g
e
d
e
tectio
n
s
h
o
ws
th
e
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
in
th
e
th
r
ee
en
v
i
r
o
n
m
e
n
tal
lig
h
tin
g
co
n
d
itio
n
a
n
d
th
e
e
d
g
es
o
f
o
b
s
tac
les
f
o
u
n
d
s
h
ar
p
i
n
th
e
ca
s
e
o
f
C
an
n
y
ed
g
e
d
ete
ctio
n
wh
er
e
as
o
th
er
ed
g
e
d
etec
tio
n
tech
n
iq
u
es
ar
e
n
o
t
a
b
le
to
d
is
p
lay
th
e
s
h
ar
p
n
ess
o
f
th
e
o
b
s
tacle
as sh
o
wn
in
Fig
u
r
e
s
7
(
a)
-
7
(
e)
(
a)
(
b
)
(
c
)
(
d
)
(
e
)
Fig
u
r
e
7
.
E
d
g
e
d
etec
tio
n
co
m
p
ar
is
o
n
o
n
en
v
ir
o
n
m
e
n
tal
lig
h
tin
g
:
(
a)
o
r
ig
in
al
im
ag
e
,
(
b
)
C
an
n
y
,
(
c)
p
r
ewith
,
(
d
)
s
o
b
el,
a
n
d
(
e
)
r
o
b
er
t e
d
g
e
d
etec
tio
n
tech
n
iq
u
es
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
555
-
1
5
6
8
1564
T
ab
le
1
.
Per
f
o
r
m
an
ce
co
m
p
a
r
is
o
n
o
f
ed
g
e
d
etec
tio
n
in
T
MS
S
.
N
o
.
P
a
r
a
me
t
e
r
s
C
a
n
n
y
P
r
e
w
i
t
h
S
o
b
e
l
R
o
b
e
r
t
1
si
z
e
S
mal
l
,
m
e
d
i
u
m
,
a
n
d
l
a
r
g
e
s
i
z
e
s
S
mal
l
,
m
e
d
i
u
m
,
a
n
d
l
a
r
g
e
s
i
z
e
s
S
mal
l
,
m
e
d
i
u
m
,
a
n
d
l
a
r
g
e
s
i
z
e
s
S
mal
l
,
m
e
d
i
u
m
,
a
n
d
l
a
r
g
e
s
i
z
e
s
2
U
n
f
a
mi
l
a
i
a
r
o
b
j
e
c
t
y
e
s
y
e
s
y
e
s
y
e
s
3
En
v
i
r
o
n
m
e
n
t
a
l
l
i
g
h
t
i
n
g
Ed
g
e
s
o
f
o
b
st
a
c
l
e
s
a
r
e
sh
a
r
p
Ed
g
e
s
o
f
o
b
st
a
c
l
e
s
a
r
e
n
o
t
s
h
a
r
p
Ed
g
e
s
o
f
o
b
st
a
c
l
e
s
a
r
e
n
o
t
s
h
a
r
p
Ed
g
e
s
o
f
o
b
st
a
c
l
e
s
a
r
e
n
o
t
s
h
a
r
p
4
M
u
l
t
i
p
l
e
d
e
t
e
c
t
i
o
n
y
e
s
y
e
s
y
e
s
No
5
A
v
e
r
a
g
e
t
r
a
f
f
i
c
d
e
n
si
t
y
3
0
.
1
2
2
8
.
4
5
2
6
.
6
1
2
5
.
8
8
6
S
h
a
r
p
n
e
ss
r
a
t
i
o
1
5
.
5
6
1
4
.
9
2
1
5
.
2
3
1
3
.
4
5
7
A
b
r
u
p
t
n
e
ss
0
.
2
5
0
.
2
3
0
.
1
8
0
.
1
4
8
S
e
n
s
i
t
i
v
i
t
y
f
a
c
t
o
r
0
.
6
5
0
.
6
1
0
.
6
4
0
.
5
6
9
P
r
o
c
e
ss
i
n
g
t
i
me
9
9
s
1
2
0
s
1
3
5
s
1
6
5
s
10
Ti
me
t
a
k
e
n
t
o
p
r
o
c
e
ss
a
f
r
a
m
e
0
.
1
5
s
0
.
1
8
s
0
.
2
5
s
0
.
2
5
s
11
A
v
e
r
a
g
e
f
r
a
me
p
e
r
sec
o
n
d
1
5
.
7
0
FPS
1
2
.
4
3
FPS
1
0
.
2
7
FPS
9
.
2
7
FPS
Als
o
,
it
is
f
o
u
n
d
th
at
all
th
e
ed
g
e
d
etec
tio
n
tech
n
iq
u
es
ar
e
c
ap
ab
le
o
f
tr
ac
in
g
th
e
m
u
ltip
le
d
etec
tio
n
.
T
h
e
av
er
ag
e
tr
af
f
ic
d
en
s
ities
o
f
ea
ch
ed
g
e
d
etec
tio
n
t
ec
h
n
iq
u
es
f
o
r
1
0
f
r
am
es
ar
e
d
eter
m
in
ed
.
T
h
e
co
m
p
ar
ativ
e
an
aly
s
is
o
f
ed
g
e
d
etec
tio
n
in
r
elatio
n
with
th
e
tr
af
f
ic
d
en
s
ity
is
p
r
esen
ted
in
Fig
u
r
e
8
.
T
r
af
f
ic
d
en
s
ity
r
ef
er
s
to
th
e
v
o
lu
m
e
o
f
v
eh
icles
p
r
esen
t
o
n
a
r
o
a
d
way
o
r
a
p
ar
ticu
lar
s
ec
tio
n
o
f
r
o
a
d
d
u
r
in
g
a
s
p
ec
if
ic
p
er
io
d
.
I
t
is
q
u
an
tifie
d
as
th
e
n
u
m
b
e
r
o
f
v
eh
icles
p
ass
in
g
a
d
esig
n
ated
p
o
i
n
t
with
in
a
g
iv
en
tim
e
in
ter
v
al
,
f
r
eq
u
e
n
tly
m
ea
s
u
r
ed
p
er
h
o
u
r
.
Fig
u
r
e
8
.
T
r
af
f
ic
d
e
n
s
ity
in
d
if
f
er
en
t e
d
g
e
d
etec
tio
n
tec
h
n
iq
u
es
Ass
es
s
in
g
tr
af
f
ic
d
e
n
s
ity
p
lay
s
a
cr
u
cial
r
o
le
in
u
n
d
er
s
tan
d
i
n
g
th
e
f
lo
w
o
f
tr
a
f
f
ic
an
d
th
e
d
eg
r
ee
o
f
co
n
g
esti
o
n
o
n
r
o
ad
way
s
.
C
an
n
y
ed
g
e
d
etec
tio
n
ca
n
esti
m
ate
th
e
ac
cu
r
ate
tr
af
f
ic
d
en
s
ity
co
m
p
ar
ed
to
o
th
er
ed
g
e
d
etec
tio
n
tech
n
iq
u
es.
T
h
e
SR
in
im
ag
e
p
r
o
ce
s
s
in
g
m
ea
s
u
r
es
h
o
w
clea
r
an
d
d
etailed
an
im
ag
e
ap
p
ea
r
s
.
It
’
s
a
way
to
u
n
d
er
s
tan
d
h
o
w
s
h
ar
p
o
r
b
lu
r
r
y
th
e
p
ictu
r
e
l
o
o
k
s
.
T
h
e
SR
is
esti
m
ated
u
s
in
g
th
e
(
1
0
)
.
=
ℎ
(
1
0
)
T
h
e
h
ig
h
er
th
e
v
alu
e
o
f
SR
i
s
d
esira
b
le
f
o
r
im
ag
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
e.
SR
g
iv
es
th
e
in
f
o
r
m
atio
n
ab
o
u
t
c
o
lo
u
r
c
o
r
r
ec
tio
n
,
n
o
is
e
r
ed
u
ctio
n
,
an
d
co
n
tr
ast
en
h
an
ce
m
en
t.
T
h
e
SR
o
f
d
if
f
er
en
t
ed
g
e
d
etec
tio
n
m
eth
o
d
s
ar
e
p
r
esen
ted
in
Fig
u
r
e
9
.
T
h
e
v
alu
e
o
f
a
b
r
u
p
tn
ess
in
d
ic
ates
r
ea
lis
m
an
d
im
ag
e
q
u
ality
.
T
h
e
e
x
ce
s
s
v
alu
e
o
f
ab
r
u
p
t
n
ess
lead
s
lo
s
e
o
f
r
ea
lity
.
T
h
e
s
elec
ted
ed
g
e
d
etec
tio
n
tech
n
iq
u
es
s
h
o
ws
m
o
d
er
ate
v
alu
e
o
f
ab
r
u
p
tn
ess
.
Ho
wev
er
,
th
er
e
ar
e
s
lig
h
t v
a
r
iatio
n
s
in
th
e
v
alu
es o
f
a
b
r
u
p
t
n
ess
illu
s
tr
at
ed
in
Fig
u
r
e
1
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.