T
E
L
KO
M
N
I
KA
T
e
lec
om
m
u
n
icat
ion
,
Com
p
u
t
i
n
g,
E
lec
t
r
on
ics
an
d
Cont
r
ol
Vol.
18
,
No.
4
,
Augus
t
2020
,
pp.
2027
~
203
4
I
S
S
N:
1693
-
6930,
a
c
c
r
e
dit
e
d
F
ir
s
t
G
r
a
de
by
Ke
me
nr
is
tekdikti
,
De
c
r
e
e
No:
21/E
/KP
T
/2018
DO
I
:
10.
12928/
T
E
L
KO
M
NI
KA
.
v18i4.
13968
2027
Jou
r
n
al
h
omepage
:
ht
tp:
//
jour
nal.
uad
.
ac
.
id/
index
.
php/T
E
L
K
OM
N
I
K
A
Gr
e
e
n
c
o
f
f
e
e
b
e
an
s
f
e
a
t
u
r
e
e
x
t
r
ac
t
or
u
si
n
g i
m
age
p
r
o
c
e
ssi
n
g
E
d
win
R.
Ar
b
oleda
1
,
Ar
n
e
l
C
.
F
aj
ar
d
o
2
,
Ruj
i
P
.
M
e
d
in
a
3
1
D
ep
ar
t
men
t
o
f
Co
m
p
u
t
er
an
d
E
l
ec
t
ro
n
i
c
s
E
n
g
i
n
eer
i
n
g
,
Co
l
l
eg
e
o
f
E
n
g
i
n
eeri
n
g
an
d
In
f
o
rmat
i
o
n
T
ec
h
n
o
l
o
g
y
,
Cav
i
t
e
St
at
e
U
n
i
v
er
s
i
t
y
,
Ph
i
l
i
p
p
i
n
es
2
Man
u
e
l
L
.
Q
u
ezo
n
U
n
i
v
er
s
i
t
y
,
P
h
i
l
i
p
p
i
n
es
3
T
ech
n
o
l
o
g
i
ca
l
In
s
t
i
t
u
t
e
o
f
t
h
e
Ph
i
l
i
p
p
i
n
es
,
Ph
i
l
i
p
p
i
n
es
Ar
t
icle
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
Aug
24
,
2019
R
e
vis
e
d
M
a
r
10
,
2020
Ac
c
e
pted
M
a
r
28
,
2020
T
h
i
s
s
t
u
d
y
o
ff
ers
a
n
o
v
e
l
s
o
l
u
t
i
o
n
t
o
d
eal
w
i
t
h
t
h
e
l
o
w
s
i
g
n
a
l
-
to
-
n
o
i
s
e
ra
t
i
o
an
d
s
l
o
w
e
x
ecu
t
i
o
n
ra
t
e
o
f
t
h
e
fi
r
s
t
d
eri
v
at
i
v
e
e
d
g
e
d
et
ec
t
i
o
n
a
l
g
o
ri
t
h
m
s
n
amel
y
,
Ro
b
ert
s
,
Prew
i
t
t
an
d
So
b
e
l
al
g
o
ri
t
h
m
s
.
Si
n
c
e
t
h
e
t
w
o
p
r
o
b
l
ems
are
b
ro
u
g
h
t
ab
o
u
t
b
y
t
h
e
c
o
mp
l
ex
ma
t
h
ema
t
i
ca
l
o
p
er
at
i
o
n
s
b
ei
n
g
u
s
e
d
b
y
t
h
e
a
l
g
o
ri
t
h
m
s
,
t
h
e
s
e
w
ere
re
p
l
a
ced
b
y
a
d
i
s
cri
m
i
n
an
t
.
T
h
e
d
e
v
el
o
p
e
d
d
i
s
cri
m
i
n
a
n
t
,
eq
u
i
v
al
e
n
t
t
o
t
h
e
p
ro
d
u
c
t
o
f
t
o
t
a
l
d
i
ffere
n
ce
an
d
i
n
t
en
s
i
t
y
d
i
v
i
d
ed
b
y
t
h
e
n
o
rma
l
i
za
t
i
o
n
v
al
u
es
,
i
s
b
as
e
d
o
n
t
h
e
“p
i
x
e
l
p
ai
r
fo
rma
t
i
o
n
”
t
h
at
p
ro
d
u
ce
s
o
p
t
i
ma
l
p
eak
s
i
g
n
al
t
o
n
o
i
s
e
rat
i
o
.
Res
u
l
t
s
o
f
t
h
e
s
t
u
d
y
ap
p
l
y
i
n
g
t
h
e
d
i
s
cri
m
i
n
a
n
t
fo
r
t
h
e
ed
g
e
d
et
ec
t
i
o
n
o
f
g
reen
co
ffee
b
ean
s
s
h
o
w
s
i
mp
r
o
v
eme
n
t
i
n
t
erms
o
f
p
ea
k
s
i
g
n
a
l
t
o
n
o
i
s
e
rat
i
o
(PSN
R),
mean
s
q
u
are
erro
r
(
MSE
),
an
d
ex
ecu
t
i
o
n
t
i
me.
It
w
as
d
et
e
rmi
n
ed
t
h
a
t
accu
racy
l
ev
e
l
v
ari
e
d
acco
rd
i
n
g
t
o
t
h
e
t
o
t
al
d
i
fferen
ce
o
f
p
i
x
el
v
a
l
u
e
s
,
i
n
t
en
s
i
t
y
,
an
d
n
o
rma
l
i
za
t
i
o
n
v
al
u
es
.
U
s
i
n
g
t
h
e
d
ev
e
l
o
p
ed
e
d
g
e
d
et
ec
t
i
o
n
t
ech
n
i
q
u
e
l
ed
t
o
i
m
p
ro
v
eme
n
t
s
i
n
t
h
e
PSN
R
o
f
2
.
0
9
1
%
,
1
.
1
6
%
,
an
d
2
.
4
7
%
o
v
er
So
b
el
,
Pre
w
i
t
t
,
an
d
Ro
b
er
t
s
re
s
p
ec
t
i
v
el
y
.
Mean
w
h
i
l
e,
i
m
p
ro
v
emen
t
i
n
t
h
e
MSE
w
a
s
meas
u
red
t
o
b
e
1
3
.
0
6
%
,
7
.
4
8
%
,
an
d
1
5
.
3
1
%
o
v
er
t
h
e
t
h
ree
a
l
g
o
ri
t
h
ms
.
L
i
k
ew
i
s
e,
i
mp
r
o
v
eme
n
t
i
n
ex
ec
u
t
i
o
n
t
i
me
w
as
a
l
s
o
ach
i
ev
e
d
at
v
al
u
es
o
f
6
9
.
0
2
%
,
6
7
.
4
0
%
,
an
d
6
5
.
4
6
%
o
v
er
S
o
b
e
l
,
Prew
i
t
t
,
an
d
Ro
b
er
t
s
r
es
p
ec
t
i
v
el
y
.
K
e
y
w
o
r
d
s
:
I
mage
pr
oc
e
s
s
ing
Ne
w
e
dge
de
tec
ti
on
a
lgor
it
hm
P
r
e
witt
R
obe
r
ts
S
obe
l
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
ce
n
s
e
.
C
or
r
e
s
pon
din
g
A
u
th
or
:
E
dwin
R
.
Ar
boleda
,
De
pa
r
tm
e
nt
of
C
omput
e
r
a
nd
E
lec
tr
onics
E
nginee
r
ing,
C
oll
e
ge
of
E
nginee
r
ing
a
nd
I
nf
or
mation
T
e
c
hnolo
gy,
C
a
vit
e
S
tate
Unive
r
s
it
y,
I
nda
ng,
C
a
vit
e
,
P
hil
ippi
ne
.
E
mail:
e
dwin
.
r
.
a
r
boleda
@c
vs
u.
e
du.
ph
1.
I
NT
RODU
C
T
I
ON
C
of
f
e
e
ha
s
a
c
hieve
d
notable
s
tatus
in
the
wor
l
d
mar
ke
t
a
nd
is
a
majo
r
in
f
luenc
e
r
on
c
ult
u
r
e
a
nd
e
c
onomy
of
many
na
ti
ons
[
1]
.
I
n
the
c
ur
r
e
nt
wor
ld
r
a
nking,
c
of
f
e
e
is
r
a
nke
d
s
e
c
ond
to
wa
ter
a
s
the
mos
t
c
ons
umed
f
ood
pr
oduc
t
[2
]
a
nd
the
s
e
c
ond
mos
t
i
n
-
de
mand
pr
oduc
t
ne
xt
only
to
pe
t
r
oleum
[3
]
.
De
s
pit
e
thi
s
,
t
he
r
e
is
not
a
s
ingl
e
,
uni
f
ied
c
of
f
e
e
be
a
n
g
r
a
din
g
s
tanda
r
d
a
c
c
e
pted
a
nd
a
dopted
by
a
ll
c
of
f
e
e
e
xpor
ti
ng
c
ountr
ies
[4
]
.
E
a
c
h
na
ti
on
e
it
he
r
f
oll
ows
a
n
e
xis
ti
ng
s
tanda
r
d
or
ha
s
c
r
e
a
ted
it
s
own
s
tanda
r
ds
[5
].
B
e
f
or
e
gr
e
e
n
c
of
f
e
e
be
a
ns
c
a
n
be
c
la
s
s
if
ied
a
nd
gr
a
de
d,
their
f
e
a
tur
e
s
mus
t
f
ir
s
t
be
e
xtr
a
c
ted
[6
]
.
M
a
ny
kinds
of
r
e
s
e
a
r
c
h
ha
ve
be
e
n
done
in
g
r
e
e
n
c
of
f
e
e
be
a
ns
f
e
a
tur
e
e
xt
r
a
c
ti
on
whic
h
include
s
the
u
s
e
o
f
c
he
mi
c
a
ls
[
7
,
8
]
,
dif
f
e
r
e
nt
types
of
s
pe
c
tr
os
c
opy
s
uc
h
a
s
f
luor
e
s
c
e
nc
e
[9
]
,
ne
a
r
inf
r
a
r
e
d
(
NI
R
S
)
[
10
,
11
]
,
F
our
ier
tr
a
ns
f
o
r
m
[
12
]
,
a
nd
R
a
man
[
13,
14
]
,
e
lec
tr
onic
tongue
[
15
]
,
e
lec
tr
onic
nos
e
(
16)
,
a
nd
im
a
ge
pr
oc
e
s
s
ing
[
17,
18
].
T
he
main
dis
a
dva
nta
ge
of
us
ing
c
he
mi
c
a
ls
f
or
f
e
a
tur
e
e
xtr
a
c
ti
on
is
the
dis
pos
a
l
of
c
he
mi
c
a
ls
a
f
ter
us
e
whic
h
a
r
e
of
tentim
e
s
ha
z
a
r
dous
to
the
e
n
vir
onment
[
19
]
.
T
he
main
dr
a
wba
c
k
of
the
us
e
o
f
dif
f
e
r
e
nt
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2027
-
2034
2028
types
of
s
pe
c
tr
os
c
opy
is
it
r
e
quir
e
s
c
os
tl
y
mac
hin
e
s
[
20
]
.
E
lec
tr
on
ic
nos
e
a
nd
e
lec
tr
onic
tongue
a
ls
o
r
e
qu
ir
e
c
ompl
ica
ted
s
e
ns
or
s
that
a
r
e
not
ve
r
y
e
a
s
y
to
s
e
t
-
up
a
nd
us
e
d
[
21
-
23
].
As
ide
f
r
om
thes
e
major
dis
a
dv
a
ntage
s
,
the
a
bove
methods
a
ls
o
ne
e
d
to
de
s
tr
oy
the
be
a
ns
a
s
they
r
e
quir
e
g
r
in
ding
,
r
e
qui
r
e
r
igo
r
ous
s
a
mpl
e
pr
e
pa
r
a
ti
on,
a
nd
ha
s
lengthy
f
e
a
tur
e
e
xtr
a
c
ti
on
in
ter
ms
of
ti
me
[
24
]
.
A
notable
e
xc
e
pti
on
is
the
us
e
of
ne
a
r
-
inf
r
a
r
e
d
s
pe
c
tr
os
c
opy
whic
h
doe
s
not
ne
e
d
to
de
s
tr
oy
s
a
mp
les
,
r
e
quir
e
s
ve
r
y
s
im
ple
s
a
mpl
e
p
r
e
pa
r
a
ti
on,
a
nd
ha
s
s
hor
t
s
pa
n
of
ti
me
f
or
f
e
a
tur
e
e
xt
r
a
c
ti
on.
Ne
ve
r
thele
s
s
,
NI
R
S
mac
hines
a
r
e
e
xpe
ns
ive
[
25
].
I
mage
pr
oc
e
s
s
ing
is
one
of
the
mos
t
pr
omi
s
ing
tec
hniques
to
e
xtr
a
c
t
gr
e
e
n
c
of
f
e
e
be
a
n
f
e
a
tur
e
s
be
c
a
us
e
it
doe
s
not
r
e
quir
e
c
ompl
ica
ted
s
e
ns
or
s
a
nd
high
ly
tec
hnica
l
e
lec
tr
onics
[
26
,
27
]
.
I
t
on
ly
ne
e
ds
a
c
a
mer
a
,
a
li
ghti
ng
mec
ha
nis
m,
a
c
omput
e
r
or
a
mi
c
r
oc
ontr
oll
e
r
a
nd
a
good
a
lgor
it
hm
f
o
r
f
e
a
tur
e
e
xtr
a
c
ti
on
[
28,
29
]
.
A
pr
incipa
l
ope
r
a
ti
on
in
im
a
ge
pr
oc
e
s
s
ing
is
the
de
tec
ti
on
of
e
dge
s
[
30
]
.
E
dge
s
a
r
e
the
b
or
de
r
li
ne
i
n
the
im
a
ge
r
e
ve
a
led
by
p
ixels
e
xhibi
ti
ng
dis
r
upti
on
in
g
r
e
y
leve
l
to
a
djoi
ning
p
ixels
[
31
]
.
T
o
s
us
tain
the
a
r
r
a
nge
ment
of
a
n
objec
t
while
a
voidi
ng
los
s
of
inf
or
mat
ion
f
or
a
nother
im
a
ge
pr
oc
e
s
s
ing
is
the
pr
im
a
r
y
objec
ti
ve
of
us
ing
e
dge
de
tec
ti
on
[
32
].
T
he
de
tec
ti
on
of
the
g
r
e
e
n
c
of
f
e
e
f
e
a
tur
e
s
us
ing
im
a
ge
pr
oc
e
s
s
ing
r
e
quir
e
e
dge
de
tec
ti
on
a
lgo
r
it
hms
.
T
his
c
a
n
be
done
us
ing
a
va
r
iety
o
f
tec
hniques
of
whic
h
f
i
r
s
t
-
or
de
r
de
r
ivative
-
ba
s
e
d
a
lgor
it
h
ms
s
uc
h
a
s
S
obe
l,
P
r
e
witt
,
a
nd
R
obe
r
ts
a
r
e
the
s
im
ples
t
a
nd
mos
t
c
omm
only
us
e
d.
Ho
we
ve
r
,
thes
e
a
l
gor
it
hms
a
r
e
s
us
c
e
pti
ble
to
nois
e
due
to
diver
s
e
modes
of
dif
f
e
r
e
nti
a
l
ope
r
a
ti
ons
[
33,
34
]
.
At
the
s
a
me
ti
me,
the
c
omput
a
ti
ona
l
c
ompl
e
xit
y
of
the
thr
e
e
a
lgor
i
thm
s
ha
s
a
ls
o
r
e
s
ult
e
d
to
lar
ge
e
xe
c
uti
on
ti
mes
[
35
-
37
]
.
As
s
uc
h,
ther
e
is
a
ne
e
d
to
a
dd
r
e
s
s
the
s
us
c
e
pti
bil
it
y
to
nois
e
of
the
f
ir
s
t
-
or
de
r
de
r
ivative
-
ba
s
e
d
a
lgor
it
hms
a
s
we
ll
a
s
their
s
low
e
xe
c
uti
on
ti
mes
a
nd
tes
t
the
ir
a
ppli
c
a
bil
it
y
towa
r
ds
the
e
dge
de
tec
ti
on
a
n
d
f
e
a
tur
e
e
xtr
a
c
ti
on
of
g
r
e
e
n
c
of
f
e
e
be
a
ns
.
2.
RE
S
E
AR
CH
M
E
T
HO
DS
2.
1
.
De
s
ign
c
on
s
id
e
r
at
i
on
s
T
he
major
c
ons
ider
a
ti
ons
f
or
the
s
tudy
a
r
e
the
t
ype
of
ha
r
dwa
r
e
a
nd
s
of
twa
r
e
to
us
e
to
a
c
hie
ve
the
objec
ti
ve
s
of
the
s
tudy.
I
t
wa
s
de
c
ided
that
the
s
uit
a
ble
s
of
twa
r
e
to
us
e
is
the
P
ython
be
c
a
us
e
of
t
he
Ope
n
C
V
whic
h
ha
s
many
li
b
r
a
r
ies
in
im
a
ge
p
r
oc
e
s
s
ing.
S
ince
the
R
a
s
pbe
r
r
y
P
i
us
e
s
P
ython
,
it
wa
s
the
ha
r
dwa
r
e
of
c
hoice
.
T
he
R
a
s
pbe
r
r
y
P
i
c
a
mer
a
modul
e
w
a
s
not
us
e
d
be
c
a
us
e
a
n
or
dinar
y
we
bc
a
m
c
a
n
be
us
e
d
f
or
the
s
a
me
r
e
s
u
lt
of
the
R
a
s
pbe
r
r
y
P
i
c
a
mer
a
.
T
he
c
os
t
wa
s
a
ls
o
c
on
s
ider
e
d
a
s
the
A4
tec
h
we
bc
a
m
is
ha
lf
the
pr
ice
of
a
R
a
s
pbe
r
r
y
P
i
C
a
mer
a
.
2.
2.
T
h
e
d
e
ve
lop
e
d
n
e
w
e
d
ge
d
e
t
e
c
t
io
n
algorit
h
m
I
n
thi
s
s
tudy,
the
dif
f
e
r
e
nc
e
s
be
twe
e
n
pixels
s
ur
r
ounding
the
pixel
of
int
e
r
e
s
t
a
s
s
hown
in
F
igur
e
1
wa
s
c
omput
e
d.
T
he
pixel
of
in
ter
e
s
t
is
the
pixel
in
the
mi
ddle
or
the
pixel
tha
t
is
s
ur
r
ounde
d
by
8
othe
r
pixels
,
na
mely
the
uppe
r
lef
t,
le
f
t
,
lowe
r
lef
t,
top
,
bott
o
m,
uppe
r
r
ight
,
r
ight
a
nd
lowe
r
r
igh
t
.
T
o
c
a
lcula
te
the
tot
a
l
dif
f
e
r
e
nc
e
,
(
1
)
wa
s
us
e
d.
T
otal
Dif
f
e
r
e
nc
e
=
(
T
op
P
ixel
Va
lue
-
B
ott
om
P
ixel
Va
lue)
+
(
1)
(
Uppe
r
L
e
f
t
P
ixel
Va
lue
-
L
owe
r
R
ight
P
ixel
Va
lue)
+
(
L
e
f
t
P
ixel
Va
lue
-
R
ight
P
ixel
Va
lue)
+
(
L
owe
r
lef
t
P
ixel
Va
lue
-
Uppe
r
R
ight
P
ixel
va
lue)
T
he
s
a
mpl
e
c
omput
a
ti
on
of
tot
a
l
dif
f
e
r
e
nc
e
whe
r
e
the
pixel
of
int
e
r
e
s
t
is
a
n
e
dge
is
s
hown
in
F
igur
e
2.
Us
ing
f
or
mul
a
number
1,
a
ll
va
lues
of
the
to
tal
di
f
f
e
r
e
nc
e
g
r
e
a
ter
than
z
e
r
o
is
c
ons
ider
e
d
a
s
a
n
e
dg
e
,
while
non
-
e
dge
will
ha
ve
a
va
lue
of
z
e
r
o.
T
he
dis
c
r
im
inant
f
or
the
ne
w
e
dge
de
tec
ti
on
a
lg
or
it
hm
is
the
tot
a
l
di
f
f
e
r
e
nc
e
mul
ti
pl
ied
by
int
e
ns
it
y
a
nd
divi
de
d
by
nor
maliza
ti
on
va
lue
a
s
s
hown
in
(
2
)
.
Ne
w
Edge
De
te
ctio
n
Dis
cr
im
in
a
nt
=
T
o
ta
l
Dif
fe
r
e
nce
∗
I
nt
e
nsi
t
y
N
or
m
a
li
za
t
i
on
(
2)
T
he
ps
e
udoc
ode
f
or
e
dge
de
tec
ti
on
is
s
hown
in
T
a
ble
1.
Us
ing
the
pr
ogr
a
m
s
hown
a
s
ps
e
udoc
ode
in
T
a
ble
1
,
e
xpe
r
im
e
nts
we
r
e
made
to
de
ter
mi
ne
whic
h
pos
s
ibl
e
c
ombi
na
ti
ons
of
pixel
dif
f
e
r
e
nc
e
s
be
twe
e
n
pixel
va
lues
of
top
a
nd
bott
om,
l
e
f
t
a
nd
r
ight
,
to
p_lef
t
a
nd
top_r
ight
bo
tt
om_l
e
f
t
a
nd
bo
tt
om_r
igh
t
yielde
d
the
be
s
t
e
dge
de
tec
ti
on.
T
he
be
s
t
c
ombi
na
ti
on
ba
s
e
d
on
the
output
pr
oduc
e
d
is
the
tot
a
l
dif
f
e
r
e
nc
e
of
t
op
pixel
mi
nus
bott
om
pixel
plus
le
f
t
pixel
mi
nus
r
ight
pixe
l
that
is
the
r
e
a
s
on
why
other
dif
f
e
r
e
nc
e
s
we
r
e
not
i
nc
luded
in
the
ps
e
udoc
ode
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
Gr
e
e
n
c
off
e
e
be
ans
featur
e
e
x
tr
a
c
tor
us
ing
image
pr
oc
e
s
s
ing
(
E
dw
in
R
.
A
r
boleda
)
2029
F
igur
e
1.
Dif
f
e
r
e
nc
e
s
of
pixel
va
lue
wa
s
c
omput
e
d
to
c
las
s
if
y
pixel
o
f
int
e
r
e
s
t
F
igur
e
2.
S
a
mpl
e
c
ompu
t
a
ti
on
of
the
to
tal
dif
f
e
r
e
n
c
e
T
a
ble
1.
T
he
ps
e
udoc
ode
f
o
r
the
ne
w
e
dge
de
tec
ti
o
n
a
lgor
it
hm
Algorithm
# program start
Import OpenCV library
start
# input section
define the normal value for x
x value to be divided by 500
define the new edge algorithm parameters
-
image and intensity
define height I and width j of an image
define the edge
# processing section
for all height i and width j pixels in range
extract pixel values of
top and bottom
left and right
top_left and top_right
bottom_left and bottom_right
extract differences
difference 1 = top minus bottom
difference 2 = left minus right
extract total difference
total difference = difference 1 + difference 2
total
difference = normal (total difference) * intensity
extract pixels of the image
image_pix = image [i , j]
extract edge_image
edge_image [ i , j] = image_pix * total difference
# output
section
display input image
display input image converted to gray scale
display detected edge
end
#eop
A
nor
malizing
va
lue
is
a
r
a
ti
o,
whic
h
mea
ns
that
f
or
a
ny
va
lue
of
tot
a
l
dif
f
e
r
e
nc
e
,
it
is
divi
de
d
by
500.
T
he
nor
malize
d
va
lue
is
then
mul
ti
pli
e
d
to
the
int
e
ns
it
y
whic
h
wa
s
given
a
v
a
lue
of
10.
W
it
hout
the
nor
malizing
va
lue
a
nd
int
e
ns
it
y
va
lue,
the
e
dge
that
wa
s
pr
oduc
e
d
by
us
ing
only
the
tot
a
l
di
f
f
e
r
e
n
c
e
is
not
c
onti
nuous
.
I
n
the
ps
e
udoc
ode
s
hown
in
T
a
ble
1,
the
nor
malize
d
va
lue
is
1/500
a
nd
x
is
the
tot
a
l
dif
f
e
r
e
nc
e
.
T
he
pr
ogr
a
m
f
lowc
ha
r
t
of
the
s
ys
tem
is
s
hown
in
F
igur
e
3.
T
he
gr
e
e
n
c
of
f
e
e
be
a
ns
s
a
mpl
e
s
we
r
e
a
r
r
a
nge
d
m
a
nua
ll
y
by
ha
nd
to
the
s
a
mpl
e
f
e
e
de
r
.
T
he
whole
pr
ogr
a
m
wa
s
wr
it
ten
in
P
ython
f
or
the
p
r
otot
ype
.
T
he
s
tar
t
of
the
pr
og
r
a
m
is
by
s
e
tt
ing
up
the
s
li
de
r
butt
on
s
o
that
the
us
e
r
c
a
n
s
e
t
-
up
the
Nor
malize
va
lue.
T
he
n
the
im
a
ge
is
c
a
ptur
e
d
by
th
e
c
a
mer
a
a
nd
e
dge
is
de
tec
ted
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2027
-
2034
2030
us
ing
the
ne
wly
de
ve
loped
a
lgor
it
hm
.
T
he
e
dge
is
ne
c
e
s
s
a
r
y
to
de
ter
mi
ne
if
the
objec
t
is
a
c
of
f
e
e
be
a
n
or
not
.
I
f
the
objec
t
ha
s
a
n
e
dge
,
ther
e
f
or
e
,
it
is
a
c
of
f
e
e
,
th
e
c
ontour
wi
ll
be
de
ter
mi
ne
d
a
nd
a
bound
ing
box
w
il
l
f
e
nc
e
the
c
of
f
e
e
be
a
ns
.
T
he
bounding
box
is
the
r
e
gion
o
f
int
e
r
e
s
t.
2.
3
.
T
h
e
d
e
ve
lop
e
d
h
ar
d
war
e
f
or
t
h
e
p
r
ot
ot
yp
e
T
he
us
e
r
c
a
n
a
c
c
e
s
s
the
pr
o
tot
ype
th
r
ough
th
e
gr
a
phica
l
us
e
r
int
e
r
f
a
c
e
s
hown
in
F
igur
e
4.
T
he
gr
a
phica
l
us
e
r
int
e
r
f
a
c
e
(
GU
I
)
of
the
pr
otot
ype
ha
s
5
butt
ons
c
or
r
e
s
ponding
to
dif
f
e
r
e
nt
f
unc
ti
ons
,
na
mely
nor
malize
,
c
a
ptur
e
,
e
dge
,
da
ta,
a
nd
s
a
ve
.
T
he
No
r
malize
butt
on
is
a
s
li
de
r
butt
on
whe
r
e
the
va
lue
c
a
n
be
a
djus
ted
f
r
om
0
to
100.
T
he
C
a
ptur
e
butt
on
is
f
o
r
t
a
king
the
im
a
ge
s
of
the
c
of
f
e
e
be
a
ns
.
I
n
F
igur
e
4
t
he
r
e
a
r
e
two
im
a
ge
s
c
a
ptur
e
d
by
the
pr
o
tot
ype
,
the
s
ize
s
of
th
e
s
e
im
a
ge
s
a
r
e
300x300
p
ixels
a
nd
800x6
00
pixels
r
e
s
pe
c
ti
ve
ly.
E
dge
butt
on
is
f
or
s
howing
the
e
dge
de
tec
ted
by
the
de
ve
loped
a
lgor
it
hm
.
T
he
Da
ta
but
ton
is
f
or
e
xtr
a
c
ti
ng
the
f
e
a
tur
e
s
of
the
c
of
f
e
e
be
a
ns
.
W
he
n
the
e
dge
is
de
tec
ted,
the
pr
ogr
a
m
then
de
ter
mi
ne
s
the
c
ontour
ins
ide
the
e
dge
.
A
r
e
c
tangula
r
bound
ing
box
indi
c
a
tes
that
the
e
dge
a
nd
c
ontour
ha
ve
be
e
n
de
tec
ted
a
n
d
ins
ide
the
bounding
box
the
mor
phology,
c
olor
,
a
nd
te
xtu
r
e
will
be
e
xtr
a
c
ted.
T
he
s
a
ve
butt
on
is
f
o
r
s
a
ving
the
e
xtr
a
c
ted
f
e
a
tur
e
s
in
the
USB
f
las
h
dr
ive
.
Th
e
f
il
e
s
c
a
n
be
view
e
d
us
ing
the
L
ibr
e
Of
f
ice
C
a
lc
in
the
s
pr
e
a
ds
he
e
t
f
or
mat
a
s
s
hown
in
F
igur
e
5.
F
igur
e
3.
T
he
pr
ogr
a
m
f
lowc
ha
r
t
f
or
the
s
ys
tem
F
igur
e
4.
T
he
gr
a
phica
l
us
e
r
int
e
r
f
a
c
e
F
igur
e
5.
T
he
e
xtr
a
c
ted
f
e
a
tur
e
s
view
e
d
us
ing
the
L
ibr
e
Of
f
ice
C
a
lc
T
he
f
ir
s
t
c
olum
n
of
the
L
ibr
e
C
a
lc
Of
f
ice
s
pr
e
a
ds
he
e
t
s
hown
in
F
igur
e
5
is
the
s
a
mpl
e
num
be
r
of
the
c
of
f
e
e
be
a
n
a
nd
the
r
e
s
t
of
the
c
olu
mns
a
r
e
t
he
21
e
xtr
a
c
ted
f
e
a
tur
e
s
f
r
om
that
be
a
n.
T
he
bott
o
m
view
of
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
Gr
e
e
n
c
off
e
e
be
ans
featur
e
e
x
tr
a
c
tor
us
ing
image
pr
oc
e
s
s
ing
(
E
dw
in
R
.
A
r
boleda
)
2031
the
pr
otot
ype
is
s
hown
in
F
igur
e
6
.
T
he
L
E
D
s
t
r
ip
powe
r
e
d
by
12
V
DC
a
da
p
ter
s
e
r
ve
s
a
s
the
li
ghti
ng
mec
ha
nis
m
of
the
p
r
otot
ype
s
hown
in
F
igu
r
e
6
.
B
e
c
a
us
e
of
the
major
a
dva
ntage
of
the
de
ve
loped
a
l
gor
it
hm
whe
r
e
in,
the
i
mage
pa
r
a
mete
r
s
c
a
n
be
a
djus
ted
u
s
ing
the
Nor
malize
d
butt
on,
the
de
vice
c
a
n
f
unc
t
ion
e
ve
n
without
the
L
E
D
s
tr
ip
ON
a
s
long
a
s
ther
e
is
a
mpl
e
li
ght
in
the
s
ur
r
ounding
a
r
e
a
or
in
the
r
oom.
I
n
t
he
c
e
nter
of
F
igur
e
6
is
the
A4te
c
h
we
bc
a
m
P
K
-
835G,
it
is
a
low
-
c
os
t
c
a
mer
a
s
uit
a
ble
f
or
the
pr
otot
ype
be
c
a
u
s
e
ther
e
is
a
n
a
djus
tm
e
nt
whe
r
e
pa
r
a
mete
r
s
of
the
im
a
ge
c
a
n
be
mani
pulate
d.
T
he
de
ve
loped
p
r
otot
ype
in
thi
s
s
tudy
is
s
hown
in
F
igur
e
7
.
F
igur
e
6.
P
r
o
tot
ype
bott
om
view
s
howing
the
L
E
D
s
tr
ip
a
nd
we
bc
a
m
F
igur
e
7.
T
he
pr
otot
ype
with
the
he
ight
a
djus
tm
e
nt
s
e
t
-
up
T
he
de
s
ign
o
f
the
p
r
otot
ype
wa
s
ba
s
e
d
on
a
mi
c
r
os
c
ope
whe
r
e
in
the
he
ight
o
f
the
c
a
mer
a
c
a
n
be
a
djus
ted.
F
or
mos
t
of
the
tes
ti
ng
done
in
the
pr
ot
otype,
the
c
a
mer
a
modul
e
s
topper
wa
s
s
e
t
-
up
a
t
14.
5
c
m.
At
thi
s
he
ight
,
the
c
a
mer
a
c
a
n
take
40
be
a
ns
a
t
a
ti
me.
T
he
li
ghti
ng
wa
s
opti
mum
a
t
thi
s
he
ight
without
pr
oduc
ing
a
ny
s
ha
dows
.
M
or
e
be
a
ns
c
a
n
be
c
a
ptur
e
d
a
t
higher
s
e
t
-
up,
but
the
a
ddit
ional
li
ghti
ng
ne
e
ds
to
be
a
dde
d.
T
he
ba
s
e
d
is
f
r
om
a
n
old
e
lec
tr
on
ic
d
r
il
l
us
e
d
f
or
dr
il
li
ng
pr
int
e
d
c
ir
c
uit
boa
r
d.
3.
RE
S
UL
T
S
AN
D
DI
S
CU
S
S
I
ON
3.
1.
T
h
e
li
s
t
of
f
e
at
u
r
e
s
t
h
at
c
an
b
e
e
x
t
r
ac
t
e
d
b
y
t
h
e
p
r
ot
ot
yp
e
M
or
e
than
2000
R
obus
ta
g
r
e
e
n
c
of
f
e
e
be
a
ns
we
r
e
us
e
d
f
or
tes
ti
ng
a
nd
e
va
luation
of
the
pr
otot
ype
.
T
he
gr
e
e
n
c
of
f
e
e
be
a
ns
a
r
e
f
r
om
I
nda
ng,
C
a
vit
e
f
r
om
whic
h
mor
phology
,
c
olor
a
nd
textur
e
f
e
a
tur
e
s
we
r
e
e
xtr
a
c
ted.
T
a
ble
2
s
hows
the
li
s
t
of
f
e
a
tur
e
s
that
c
a
n
be
e
xtr
a
c
ted
by
the
pr
ogr
a
m.
T
he
r
e
a
r
e
a
tot
a
l
of
21
f
e
a
tur
e
s
that
c
a
n
be
e
xtr
a
c
ted
by
the
pr
ogr
a
m.
I
t
is
c
om
pos
e
d
of
11
mor
pholog
ica
l
f
e
a
tur
e
s
,
6
c
olor
f
e
a
tur
e
s
a
nd
4
textur
e
f
e
a
tur
e
s
.
3
.
2
.
C
o
m
p
a
r
i
s
o
n
o
f
t
h
e
d
e
v
e
l
o
p
e
d
a
l
g
o
r
i
t
h
m
w
i
t
h
t
h
e
S
o
b
e
l
,
P
r
e
w
i
t
t
,
a
n
d
R
o
b
e
r
t
s
e
d
g
e
d
e
t
e
c
t
i
o
n
a
l
g
o
r
i
t
h
m
T
he
de
ve
loped
e
dge
de
tec
ti
on
a
lgor
it
hm
wa
s
loade
d
in
a
c
omput
e
r
a
nd
it
s
pe
r
f
or
manc
e
wa
s
c
ompar
e
d
to
S
obe
l,
P
r
e
witt
a
nd
R
obe
r
ts
E
dge
a
lgo
r
it
hm
a
s
s
hown
in
F
igu
r
e
8
.
Us
ing
vis
ua
l
obs
e
r
va
ti
on
on
F
igur
e
8
,
a
nd
it
c
a
n
be
s
e
e
n
that
the
de
ve
loped
a
lgor
it
hm
i
s
be
tt
e
r
in
de
tec
ti
ng
e
dge
s
a
s
c
ompar
e
d
to
S
obe
l,
P
r
e
witt
,
a
nd
R
obe
r
ts
a
lg
or
it
hm.
T
he
i
mage
s
in
F
igur
e
8
we
r
e
c
ompar
e
d
us
ing
the
mea
n
s
qua
r
e
d
e
r
r
or
(
M
S
E
)
,
pe
a
k
s
ignal
to
nois
e
r
a
ti
o
(
P
S
NR
)
a
nd
e
xe
c
uti
on
ti
mes
.
All
the
im
a
ge
s
a
r
e
made
s
a
me
s
ize
,
be
c
a
us
e
the
P
ython
pr
ogr
a
m
f
unc
ti
ons
on
ly
if
they
a
r
e
of
dif
f
e
r
e
nt
s
ize
s
.
E
xe
c
uti
on
t
im
e
s
of
the
ne
wly
de
ve
loped
a
lgor
it
hm
a
nd
the
thr
e
e
f
ir
s
t
or
de
r
e
dge
de
tec
ti
on
a
lgo
r
it
h
ms
we
r
e
indepe
nde
ntl
y
ti
med,
mea
s
ur
e
d
a
nd
c
o
mpar
e
d.
T
he
r
e
s
ult
o
f
the
e
va
luation
is
s
hown
in
T
a
ble
3
.
T
he
de
ve
loped
a
lgor
it
hm
ha
s
lowe
r
mea
n
s
qua
r
e
e
r
r
or
c
ompar
e
d
to
S
obe
l,
P
r
e
witt
a
nd
R
obe
r
ts
e
dge
de
tec
ti
on
a
lgor
it
hm
whic
h
mea
ns
that
the
di
f
f
e
r
e
nc
e
be
twe
e
n
the
o
r
igi
na
l
im
a
ge
a
nd
the
im
a
ge
with
the
de
tec
ted
e
dge
is
lowe
s
t
a
s
c
ompar
e
d
to
the
oth
e
r
thr
e
e
a
lgor
it
hms
.
T
he
de
ve
loped
a
lgor
it
h
m
got
t
he
higher
P
S
NR
,
thi
s
c
onf
i
r
ms
the
vis
ua
l
ins
pe
c
ti
on
whic
h
m
e
a
ns
that
the
pe
a
k
s
ignal
is
higher
in
the
de
ve
loped
a
lgor
it
hm
a
s
c
ompar
e
d
to
nois
e
.
Als
o,
e
xe
c
uti
on
ti
mes
in
th
e
de
ve
loped
a
lgor
it
hm
is
f
a
s
ter
than
a
ll
thr
e
e
f
ir
s
t
de
gr
e
e
de
r
ivative
a
lgor
it
hms
.
B
e
tt
e
r
pe
r
f
or
manc
e
of
the
de
ve
loped
a
lgor
i
thm
a
s
c
ompar
e
d
to
other
thr
e
e
c
onve
nti
ona
l
a
lgo
r
it
hms
c
a
n
be
a
tt
r
ibut
e
d
to
the
a
djus
tm
e
nts
done
by
nor
maliza
ti
on
to
the
im
a
ge
.
I
n
S
obe
l,
P
r
e
witt
a
nd
R
obe
r
ts
e
dge
de
tec
ti
on
s
hown
in
F
igur
e
8,
the
nois
e
ins
ide
the
e
dge
of
the
c
of
f
e
e
be
a
ns
c
a
n
be
c
lea
r
ly
s
e
e
n,
whe
r
e
a
s
in
us
ing
the
ne
wly
de
ve
loped
e
dge
de
tec
ti
on
thes
e
nois
e
s
we
r
e
e
li
mi
na
ted.
Als
o,
the
e
xe
c
uti
on
ti
me
of
the
ne
wly
de
ve
loped
a
lgor
it
hm
is
2
.
49
s
e
c
onds
whic
h
is
f
a
s
ter
than
a
ll
th
r
e
e
c
onve
nti
ona
l
f
ir
s
t
de
gr
e
e
de
r
ivative
a
lgor
it
hms
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2027
-
2034
2032
T
a
ble
2.
F
e
a
tur
e
s
that
c
a
n
be
e
xtr
a
c
ted
by
the
pr
og
r
a
m
M
or
phol
ogy
C
ol
or
T
e
xt
ur
e
A
r
e
a
A
ve
r
a
ge
R
e
d
E
nt
r
opy
P
e
r
im
e
te
r
A
ve
r
a
ge
B
lu
e
C
ont
r
a
s
t
L
e
ngt
h
A
ve
r
a
ge
G
r
e
e
n
E
ne
r
gy
W
id
th
A
ve
r
a
ge
H
ue
H
omoge
ne
it
y
R
e
c
ta
ngul
a
r
A
s
p
e
c
t
R
a
ti
o
A
ve
r
a
ge
S
a
tu
r
a
ti
on
A
s
pe
c
t
R
a
ti
o
A
ve
r
a
ge
V
a
lu
e
(
B
r
ig
ht
ne
s
s
)
M
a
jo
r
A
xi
s
M
in
or
A
xi
s
R
oundne
s
s
F
e
r
r
e
t
D
ia
me
te
r
E
c
c
e
nt
r
ic
it
y
T
a
ble
3.
T
he
de
ve
loped
a
lgo
r
it
hm
Vs
.
S
obe
l
Vs
.
P
r
e
witt
Vs
.
R
obe
r
ts
E
dge
D
e
te
c
to
r
M
S
E
P
S
N
R
E
xe
c
ut
io
n T
im
e
s
(
S
)
T
hi
s
s
tu
dy
68. 55
29.77
2.49
S
obe
l
78. 85
29.16
8.04
P
r
e
w
it
t
74. 10
29.43
7.64
R
obe
r
ts
80.95
29.05
7.21
(
a
)
(
b)
(
c
)
(
d)
(
e
)
F
igur
e
8.
Or
igi
na
l
im
a
ge
a
nd
the
dif
f
e
r
e
nt
e
dge
s
de
r
ive
us
ing
dif
f
e
r
e
nt
a
lgo
r
it
hms
;
(
a
)
or
igi
na
l
im
a
ge
,
(
b)
ne
wly
d
e
ve
loped
e
dge
de
tec
ti
on
a
lgor
it
hm
,
(
c
)
S
obe
l
e
dge
de
tec
ti
on
a
lgor
it
h
m,
(
d)
P
r
e
witt
e
dge
de
tec
ti
on
a
lgor
it
hm
,
(
e
)
R
obe
r
ts
e
dge
de
tec
ti
on
a
lgor
it
hm
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
Gr
e
e
n
c
off
e
e
be
ans
featur
e
e
x
tr
a
c
tor
us
ing
image
pr
oc
e
s
s
ing
(
E
dw
in
R
.
A
r
boleda
)
2033
4.
CONC
L
USI
ONS
B
a
s
e
d
on
the
r
e
s
ult
s
of
the
di
f
f
e
r
e
nt
tes
ts
c
onduc
ted
i
n
thi
s
s
tudy,
the
f
oll
owing
c
onc
lus
ions
a
r
e
d
r
a
wn:
A
dis
c
r
im
inant
f
o
r
e
dge
de
tec
ti
on
wa
s
de
ve
loped
a
s
de
f
ined
a
s
Ne
w
Edge
De
te
ctio
n
Dis
cr
im
in
a
nt
=
T
o
ta
l
Dif
fe
r
e
nce
∗
I
nt
e
ns
it
y
No
r
m
a
l
iz
a
ti
o
n
As
a
ppli
e
d
to
c
of
f
e
e
be
a
ns
,
the
opti
mal
va
lue
s
a
r
e
10
f
or
int
e
ns
it
y;
a
nd
25
to
40
f
or
nor
m
a
li
z
a
ti
on
a
nd
de
pe
nde
nt
on
im
a
ge
.
T
he
pixel
pa
ir
f
or
mation
that
pr
oduc
e
s
opti
mal
pe
a
k
s
ignal
to
nois
e
r
a
ti
o
is
the
s
um
of
the
di
f
f
e
r
e
nc
e
o
f
top
pixel
a
nd
bott
om
pixel
a
nd
dif
f
e
r
e
nc
e
of
lef
t
pixel
a
nd
r
ight
pixel.
T
he
im
p
r
ove
ment
of
the
de
ve
loped
e
dge
de
tec
ti
on
a
lgor
it
hm
ove
r
S
obe
l,
P
r
e
wit
t,
a
nd
R
obe
r
ts
a
r
e
:
2.
091
%
,
1
.
16%
,
a
nd
2
.
47%
in
ter
ms
of
P
S
NR
,
r
e
s
pe
c
ti
ve
ly;
13.
06%
,
7.
48
%
,
a
nd
15.
31%
in
ter
ms
of
M
S
E
,
r
e
s
pe
c
ti
ve
ly;
69.
02%
,
6
7.
40%
,
a
nd
65.
46%
in
ter
ms
o
f
e
xe
c
uti
on
ti
me
,
r
e
s
pe
c
ti
ve
ly.
T
he
c
of
f
e
e
be
a
n
f
e
a
tur
e
s
e
xtr
a
c
ted
by
the
pr
otot
ype
a
r
e
:
a
r
e
a
,
pe
r
i
mete
r
,
length,
width,
r
e
c
tangula
r
a
s
pe
c
t
r
a
ti
o,
a
s
pe
c
t
r
a
ti
o,
major
a
xis
,
mi
nor
a
xis
,
r
oundne
s
s
,
f
e
r
r
e
t
diame
ter
,
e
c
c
e
ntr
icity
f
or
mo
r
phol
ogy;
a
ve
r
a
ge
r
e
d,
a
ve
r
a
ge
b
lue,
a
ve
r
a
ge
g
r
e
e
n,
a
ve
r
a
ge
hue
,
a
ve
r
a
ge
s
a
tur
a
ti
on,
a
ve
r
a
ge
va
lues
f
or
c
olor
;
a
nd
e
ntr
opy,
c
ont
r
a
s
t,
e
ne
r
gy
,
a
nd
ho
mogene
it
y
f
o
r
text
ur
e
.
Ove
r
a
ll
,
i
t
c
a
n
be
c
laimed
that
the
ge
ne
r
a
l
objec
ti
ve
of
thi
s
s
tudy
whic
h
is
to
de
ve
lop
a
ne
w
e
dge
de
tec
ti
on
a
ppr
oa
c
h
f
or
gr
e
e
n
c
o
f
f
e
e
be
a
ns
f
e
a
tur
e
h
a
s
be
e
n
a
c
hieve
d.
T
he
main
objec
ti
ve
ha
s
be
e
n
a
c
hi
e
ve
d
by
us
ing
the
he
ur
is
ti
c
a
ppr
oa
c
h
in
c
a
lcula
ti
ng
the
r
ight
va
lues
f
or
the
dis
c
r
im
inant
a
nd
f
indi
ng
the
b
e
s
t
pixel
f
or
mation
that
p
r
oduc
e
s
the
op
ti
mal
P
S
NR
.
T
he
ne
wly
de
ve
loped
e
dge
de
tec
ti
on
a
lgor
it
h
m
wa
s
p
r
ove
n
be
tt
e
r
than
R
obe
r
ts
,
P
r
e
witt
a
nd
S
obe
l
a
lgor
i
thm
in
ter
ms
of
P
S
NR
,
M
S
E
a
nd
e
xe
c
uti
on
ti
mes
.
Upon
in
tegr
a
ti
ng
the
ne
wly
de
ve
loped
a
lgo
r
it
hm
in
a
R
a
s
pbe
r
r
y
P
i
mi
c
r
oc
omput
e
r
,
the
de
ve
loped
ha
r
dwa
r
e
wa
s
a
ble
to
e
xtr
a
c
t
the
mor
phologi
c
a
l,
c
olor
a
nd
textur
e
f
e
a
tu
r
e
s
of
th
e
gr
e
e
n
c
of
f
e
e
be
a
ns
.
RE
F
E
RE
NC
E
S
[1
]
Cao
E
.
P
.
,
Co
n
s
t
an
t
i
n
o
-
Sa
n
t
o
s
D
.
M
.
,
A
n
j
e
l
a
L
.
,
Ram
o
s
P
.
,
San
t
o
s
B
.
S
.
,
Q
u
i
l
an
g
J
.
P
.
,
et
al
.
,
"
Mo
l
ec
u
l
ar
an
d
mo
rp
h
o
l
o
g
i
ca
l
d
i
fferen
t
i
a
t
i
o
n
am
o
n
g
C
o
ffea
(Ru
b
i
ac
eae)
v
ari
e
t
i
e
s
g
r
o
w
n
i
n
t
h
e
farm
s
o
f
Cav
i
t
e
Pr
o
v
i
n
ce
,
Ph
i
l
i
p
p
i
n
es
,
"
P
h
i
l
i
p
p
S
ci
Let
t
,
V
o
l
.
3
8
7
,
N
o
.
2
,
2
0
1
4
.
[ci
t
ed
2
0
1
7
J
u
n
3
]
[In
t
ern
e
t
].
A
v
a
i
l
a
b
l
e
fr
o
m
:
h
t
t
p
:
/
/
p
h
i
l
s
ci
l
e
t
t
er
s
.
o
rg
/
2
0
1
4
/
PSL
2
0
1
4
-
v
o
l
0
7
-
n
o
0
2
-
p
3
8
7
-
3
9
7
Cao
.
p
d
f
[2
]
Farah
A
.
,
Ferrei
ra
T
.
,
"
T
h
e
Co
ffe
e
Pl
a
n
t
a
n
d
Bea
n
s
:
A
n
I
n
t
r
o
d
u
ct
i
o
n
,"
In
:
P
r
ee
d
y
V
R
,
ed
i
t
o
r
.
Co
f
f
ee
i
n
H
e
a
l
t
h
a
n
d
D
i
s
ea
s
e
P
r
even
t
i
o
n
.
A
c
a
d
e
m
i
c
P
r
es
s
.,
p
p
.
5
-
10
,
2
0
1
5
.
[3
]
N
ai
r
K
.
P
.
P.
,
"
6
-
Co
ffee
,"
In
:
Th
e
A
g
r
o
n
o
m
y
a
n
d
E
c
o
n
o
m
y
o
f
I
m
p
o
r
t
a
n
t
T
r
ee
Cr
o
p
s
o
f
t
h
e
D
eve
l
o
p
i
n
g
W
o
r
l
d
.
E
l
s
evi
e
r
,
p
p
.
1
8
1
-
2
0
8
,
2
0
1
0
.
[4
]
D
at
o
v
A
.
,
L
i
n
Y
.
-
C.
,
"
Cl
as
s
i
fi
ca
t
i
o
n
a
n
d
G
rad
i
n
g
o
f
G
reen
Co
ffee
Bea
n
s
i
n
A
s
i
a
,
"
2
0
1
9.
[In
t
ern
e
t
].
A
v
ai
l
ab
l
e
fr
o
m
:
h
t
t
p
:
/
/
ap
.
fft
c.
a
g
n
e
t
.
o
rg
/
ap
_
d
b
.
p
h
p
?
i
d
=
9
9
2
&
p
ri
n
t
=
1
[5
]
"
In
t
ern
a
t
i
o
n
a
l
Co
ffee
O
rg
a
n
i
za
t
i
o
n
.
G
rad
i
n
g
an
d
Cl
as
s
i
fi
cat
i
o
n
o
f
G
reen
Co
ffee
,"
2
0
1
8
[ci
t
e
d
2
0
1
8
Sep
2
3
].
[In
t
er
n
et
].
A
v
ai
l
ab
l
e
fro
m:
h
t
t
p
:
/
/
w
w
w
.
i
c
o
.
o
rg
/
p
r
o
j
ec
t
s
/
G
o
o
d
-
H
y
g
i
e
n
e
-
Pract
i
ces
/
cn
t
/
c
n
t
_
en
/
s
ec
_
3
/
d
o
cs
_
3
.
3
/
G
ra
d
i
n
g
&
cl
as
s
.
p
d
f
[6
]
Fran
ca
A
.
S
.
,
O
l
i
v
ei
ra
L
.
S
.
,
Men
d
o
n
c
J
.
C
.
F
.
,
Si
l
v
a
A
.
,
"
Ph
y
s
i
cal
a
n
d
ch
em
i
cal
a
t
t
r
i
b
u
t
e
s
o
f
d
efec
t
i
v
e
cr
u
d
e
an
d
ro
a
s
t
e
d
co
ffee
b
ea
n
s
,
"
F
o
o
d
Ch
em
.
V
o
l
.
90
,
p
p
.
89
-
94
,
2
0
0
5
.
[7
]
Barb
o
s
a
R
.
M
.
,
Bat
i
s
t
a
B
.
L
.
,
V
arri
q
u
e
R
.
M
.
,
Co
el
h
o
V
.
A
.
,
Camp
i
g
l
i
a
A
.
D
.
,
Barb
o
s
a
F.
,
"
T
h
e
u
s
e
o
f
ad
v
an
ced
ch
emo
me
t
ri
c
t
ech
n
i
q
u
e
s
an
d
t
race
e
l
emen
t
l
e
v
el
s
fo
r
c
o
n
t
r
o
l
l
i
n
g
t
h
e
au
t
h
en
t
i
c
i
t
y
o
f
o
rg
a
n
i
c
c
o
ffee
,"
F
o
o
d
R
es
In
t
.
V
o
l
.
61
,
p
p
.
2
4
6
-
2
51
,
2
0
1
4
.
[8
]
Cal
v
i
n
i
R
.
,
U
l
ri
c
i
A
.
,
Man
u
e
l
J
.
,
"
Ch
emo
me
t
ri
c
s
an
d
In
t
el
l
i
g
e
n
t
L
ab
o
ra
t
o
r
y
Sy
s
t
em
s
Pract
i
cal
c
o
mp
ar
i
s
o
n
o
f
s
p
a
rs
e
met
h
o
d
s
fo
r
cl
as
s
i
fi
cat
i
o
n
o
f
A
ra
b
i
ca
an
d
Ro
b
u
s
t
a
co
ffee
s
p
ec
i
es
u
s
i
n
g
n
ear
i
n
frare
d
h
y
p
ers
p
ect
ra
l
i
mag
i
n
g
,
"
Ch
emo
m
I
n
t
e
l
l
L
ab
S
y
s
t
,
V
o
l
.
1
4
6
,
p
p
.
5
0
3
-
5
11
,
2
0
1
5
.
[In
t
ern
e
t
].
A
v
ai
l
ab
l
e
fro
m
:
h
t
t
p
:
/
/
d
x
.
d
o
i
.
o
rg
/
1
0
.
1
0
1
6
/
j
.
ch
emo
l
ab
.
2
0
1
5
.
0
7
.
0
1
0
[9
]
Bo
ri
s
o
v
a
E
.
,
Y
u
l
i
a
M
.
,
A
s
n
a
n
i
n
g
A
.
R
.
,
Su
h
a
n
d
y
D
.
,
Y
u
l
i
a
M
.
,
Su
h
an
d
y
D
.
,
"
D
i
s
cri
m
i
n
a
t
i
o
n
o
f
s
ev
eral
I
n
d
o
n
e
s
i
a
n
s
p
ec
i
al
t
y
c
o
ffees
u
s
i
n
g
Fl
u
o
re
s
cen
ce
S
p
ect
r
o
s
c
o
p
y
c
o
mb
i
n
e
d
w
i
t
h
SIMC
A
met
h
o
d
D
i
s
cr
i
mi
n
at
i
o
n
o
f
s
ev
eral
In
d
o
n
e
s
i
a
n
s
p
ec
i
al
t
y
co
ffee
s
u
s
i
n
g
Fl
u
o
res
ce
n
ce
Sp
ect
r
o
s
c
o
p
y
co
mb
i
n
e
d
w
i
t
h
SIMCA
met
h
o
d
,"
In
:
3
r
d
ICCh
F
S
A
2
0
1
7
.
IO
P
Co
n
f
er
e
n
ce
S
e
r
i
e
s
:
M
a
t
er
i
a
l
s
S
ci
e
n
ce
a
n
d
E
n
g
i
n
eer
i
n
g
,
p
p
.
1
-
7
,
2
0
1
8
.
[1
0
]
A
b
i
d
i
n
S
.
Z
.
,
K
ami
l
W
.
A
.
,
J
u
l
i
an
o
P.
,
"
Po
rt
a
b
l
e
n
ear
i
n
frared
s
p
ec
t
ro
s
co
p
y
ap
p
l
i
ed
t
o
q
u
a
l
i
t
y
co
n
t
r
o
l
o
f
Braz
i
l
i
a
n
co
ffee,
"
F
o
o
d
B
i
o
s
ci
.
2
0
1
6
.
[1
1
]
Pás
co
a
R
.
N
.
M
.
J
.
,
Sarra
g
u
ça
M
.
C
.
,
Mag
a
l
h
ãe
s
L
.
M
.
,
San
t
o
s
J
.
R
.
,
Ran
g
e
l
A
.
O
.
S
.
S
.
,
L
o
p
es
J
.
A.
,
"
U
s
e
o
f
N
ear
-
In
frare
d
Sp
ec
t
ro
s
co
p
y
f
o
r
Co
ffee
Bea
n
s
Q
u
a
l
i
t
y
A
s
s
es
s
m
en
t
,
"
In
:
Co
f
f
ee
i
n
H
ea
l
t
h
a
n
d
D
i
s
ea
s
e
P
r
ev
en
t
i
o
n
,
p
p
.
9
3
3
-
9
4
2
,
2
0
1
5
.
[In
t
ern
e
t
]
A
v
a
i
l
a
b
l
e
fr
o
m:
h
t
t
p
:
/
/
d
x
.
d
o
i
.
o
r
g
/
1
0
.
1
0
1
6
/
B9
7
8
-
0
-
12
-
4
0
9
5
1
7
-
5
.
0
0
1
0
3
-
0
[1
2
]
Crai
g
A
.
P
.
,
Fran
ca
A
.
S
.
,
O
l
i
v
e
i
ra
L
.
S.
,
"
E
v
al
u
at
i
o
n
o
f
t
h
e
p
o
t
en
t
i
a
l
o
f
FT
IR
an
d
ch
em
o
met
r
i
cs
fo
r
s
ep
ara
t
i
o
n
b
et
w
een
d
efe
c
t
i
v
e
an
d
n
o
n
-
d
efect
i
v
e
co
ffee
s
,
"
F
o
o
d
Ch
e
m
.
V
o
l
.
1
3
2
,
N
o
.
3
,
p
p
.
1
3
6
8
-
1
3
7
4
,
2
0
1
2
.
[1
3
]
El
-
A
b
a
s
s
y
R
.
M
.
,
D
o
n
fac
k
P
.
,
Mat
ern
y
A
.
,
"
D
i
s
cr
i
mi
n
at
i
o
n
b
e
t
w
ee
n
A
ra
b
i
ca
a
n
d
R
o
b
u
s
t
a
g
ree
n
co
ffee
u
s
i
n
g
v
i
s
i
b
l
e
mi
cro
Raman
s
p
ec
t
ro
s
co
p
y
an
d
ch
em
o
met
r
i
c
an
a
l
y
s
i
s
,"
F
o
o
d
Ch
e
m
.
,
V
o
l
.
1
2
6
,
N
o
.
3
,
p
p
.
1
4
4
3
-
1
4
4
8
,
2
0
1
1
.
[1
4
]
K
ei
d
el
A
.
,
V
o
n
St
e
t
t
en
D
.
,
Ro
d
ri
g
u
e
s
C
.
,
Mág
u
as
C
.
,
H
i
l
d
e
b
ran
d
t
P.
,
"
D
i
s
cr
i
mi
n
at
i
o
n
o
f
g
r
een
arab
i
ca
an
d
Ro
b
u
s
t
a
co
f
fee
b
ean
s
b
y
Raman
s
p
ect
r
o
s
c
o
p
y
,
"
J
A
g
r
i
c
Fo
o
d
Ch
em.
,
V
o
l
.
5
8
,
N
o
.
2
1
,
p
p
.
1
1
1
8
7
-
1
1
1
92
,
2
0
1
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2027
-
2034
2034
[1
5
]
L
o
p
et
c
h
arat
K
.
,
K
u
l
a
p
i
c
h
i
t
r
F
.
,
Su
p
p
av
o
ras
a
t
i
t
I.
,
Ch
o
d
j
ar
u
s
a
w
ad
T
.
,
Ph
a
t
t
h
ara
-
A
n
ek
s
i
n
A
.
,
Prat
o
n
t
ep
S
.
,
et
al
.
,
"
Rel
at
i
o
n
s
h
i
p
b
et
w
een
o
v
era
l
l
d
i
ffere
n
ce
d
ec
i
s
i
o
n
an
d
el
ect
r
o
n
i
c
t
o
n
g
u
e
:
D
i
s
cr
i
mi
n
at
i
o
n
o
f
c
i
v
e
t
co
ffee
,"
J
F
o
o
d
E
n
g
.,
V
o
l
.
1
8
0
,
p
p
.
6
0
-
6
8
,
2
0
1
6
.
[In
t
er
n
et
]
A
v
a
i
l
a
b
l
e
fro
m:
h
t
t
p
:
/
/
d
x
.
d
o
i
.
o
r
g
/
1
0
.
1
0
1
6
/
j
.
j
f
o
o
d
en
g
.
2
0
1
6
.
0
2
.
0
1
1
[1
6
]
O
n
g
o
E
.
,
Fal
as
c
o
n
i
M
.
,
Sb
erv
eg
l
i
er
i
G
.
,
A
n
t
o
n
el
l
i
A
.
,
M
o
n
t
ev
ecc
h
i
G
.
,
"
Ch
emo
met
ri
c
D
i
s
cr
i
mi
n
at
i
o
n
o
f
Ph
i
l
i
p
p
i
n
e
Ci
v
e
t
Co
ffee
U
s
i
n
g
E
l
ec
t
ro
n
i
c
N
o
s
e
a
n
d
G
as
C
h
ro
ma
t
o
g
rap
h
y
Mas
s
Sp
ec
t
ro
me
t
ry
,
"
In
:
P
r
o
ced
i
a
E
n
g
i
n
eer
i
n
g
.
E
l
s
evi
e
r
B
.
V
.
,
p
p
.
9
7
7
-
9
80
,
2
0
12
.
[In
t
ern
e
t
]
A
v
ai
l
a
b
l
e
fro
m:
h
t
t
p
:
/
/
d
x
.
d
o
i
.
o
r
g
/
1
0
.
1
0
1
6
/
j
.
p
r
o
en
g
.
2
0
1
2
.
0
9
.
3
1
0
[1
7
]
H
en
d
raw
a
n
Y
.
,
W
i
d
y
an
i
n
g
t
y
as
S
.
,
Su
ci
p
t
o
S.
,
"
Co
mp
u
t
er
v
i
s
i
o
n
fo
r
p
u
r
i
t
y
,
p
h
en
o
l
,
an
d
p
H
d
et
ec
t
i
o
n
o
f
L
u
w
ak
Co
ffee
g
reen
b
ean
,"
TE
LKO
M
NIKA
Tel
eco
m
m
u
n
i
ca
t
i
o
n
Co
m
p
u
t
i
n
g
E
l
ect
r
o
n
i
c
s
a
n
d
Co
n
t
r
o
l
,
V
o
l
.
17
,
N
o
.
6
,
p
p
.
3
0
7
3
-
30
85
,
2
0
1
9
.
[1
8
]
Men
g
i
s
t
u
A
.
D.
,
"
T
h
e
E
ffec
t
s
o
f
Seg
men
t
at
i
o
n
T
ec
h
n
i
q
u
es
i
n
D
i
g
i
t
al
Imag
e
Ba
s
ed
Id
e
n
t
i
fi
ca
t
i
o
n
o
f
E
t
h
i
o
p
i
an
C
o
ffee
V
ari
e
t
y
,"
TE
LKO
M
NIK
A
Tel
eco
m
m
u
n
i
ca
t
i
o
n
Co
m
p
u
t
i
n
g
E
l
ect
r
o
n
i
c
s
a
n
d
Co
n
t
r
o
l
,
V
o
l
.
1
6
,
N
o
.
2
,
p
p
.
7
1
3
-
7
2
1
,
2
0
1
8
.
[1
9
]
W
i
l
l
i
ams
P.
,
"
N
ear
-
In
frared
T
ech
n
o
l
o
g
y
:
G
et
t
i
n
g
t
h
e
Bes
t
O
u
t
o
f
L
i
g
h
t
.
L
aw
ren
ce
S,
W
ar
b
u
r
t
o
n
P,
ed
i
t
o
rs
.
,"
V
al
u
e
A
d
d
ed
W
h
ea
t
CRC,
L
i
mi
t
ed
,
2
0
0
4
.
[2
0
]
W
i
l
l
i
ams
P.
,
"
N
ear
-
I
n
frared
T
ech
n
o
l
o
g
y
:
I
n
t
h
e
A
g
r
i
c
u
l
t
u
ra
l
an
d
Fo
o
d
In
d
u
s
t
r
i
es
2
n
d
E
d
i
t
i
o
n
.
2
n
d
e
d
.
N
o
rri
s
K
,
ed
i
t
o
r.
,"
A
meri
ca
n
A
s
s
o
c
i
at
i
o
n
o
f
Cereal
C
h
emi
s
t
s
;
2
0
0
1
.
[2
1
]
D
o
n
g
W
.
,
Z
h
ao
J
.
,
H
u
R
.
,
D
o
n
g
Y
.
,
T
an
L
.
,
"
D
i
fferen
t
i
a
t
i
o
n
o
f
Ch
i
n
es
e
ro
b
u
s
t
a
co
ffees
acco
r
d
i
n
g
t
o
s
p
eci
e
s
,
u
s
i
n
g
a
co
m
-
b
i
n
ed
e
l
ect
r
o
n
i
c
n
o
s
e
an
d
t
o
n
g
u
e,
w
i
t
h
t
h
e
a
i
d
o
f
ch
emo
me
t
ri
c
s
,
"
F
o
o
d
Ch
e
m
.
2
0
1
7
.
[2
2
]
G
u
n
aw
a
n
B
.
,
A
l
far
i
s
i
S
.
,
Sat
ri
o
G
.
,
Su
d
arma
j
i
A
.
,
"
MO
S
g
a
s
s
e
n
s
o
r
o
f
mea
t
fre
s
h
n
es
s
an
a
l
y
s
i
s
o
n
E
-
n
o
s
e
,"
TE
LKO
M
NIK
A
Tel
ec
o
m
m
u
n
i
c
a
t
i
o
n
Co
m
p
u
t
i
n
g
E
l
ec
t
r
o
n
i
c
s
a
n
d
Co
n
t
r
o
l
,
V
o
l
.
1
7
,
N
o
.
2
,
p
p
.
7
7
1
-
7
8
0
,
2
0
1
9
.
[2
3
]
Sarn
o
R
.
,
W
i
j
a
y
a
D
.
R.
,
"
Recen
t
d
ev
e
l
o
p
men
t
i
n
el
e
c
t
ro
n
i
c
n
o
s
e
d
a
t
a
p
r
o
ces
s
i
n
g
f
o
r
b
eef
q
u
al
i
t
y
as
s
es
s
me
n
t
,"
TE
LKO
M
NIK
A
Tel
ec
o
m
m
u
n
i
c
a
t
i
o
n
Co
m
p
u
t
i
n
g
E
l
ec
t
r
o
n
i
c
s
a
n
d
Co
n
t
r
o
l
,
V
o
l
.
17
,
N
o
.
1
,
p
p
.
3
3
7
-
3
48
,
2
0
1
9
.
[2
4
]
O
n
g
o
E
.
,
Fa
l
as
co
n
i
M
.
,
Sb
erv
eg
l
i
eri
G
.
,
A
n
t
o
n
e
l
l
i
A
.
,
Mo
n
t
ev
ecc
h
i
G
.
,
Sb
erv
eg
l
i
er
i
V
.
,
et
al
.
,
"
Ch
emo
met
ri
c
d
i
s
cri
m
i
n
a
t
i
o
n
o
f
p
h
i
l
i
p
p
i
n
e
ci
v
et
co
ffee
u
s
i
n
g
e
l
ect
r
o
n
i
c
n
o
s
e
an
d
g
as
c
h
ro
ma
t
o
g
rap
h
y
mas
s
s
p
ect
r
o
met
r
y
,
"
P
r
o
ce
d
i
a
E
n
g
.
,
V
o
l
.
47
,
N
o
.
9
,
p
p
.
9
77
-
9
80
,
2
0
1
2
.
[2
5
]
A
rb
o
l
e
d
a
E
.
R.
,
"
D
i
s
cri
m
i
n
a
t
i
o
n
o
f
ci
v
e
t
co
ffee
u
s
i
n
g
n
ear
i
n
frare
d
s
p
ect
r
o
s
c
o
p
y
an
d
art
i
fi
c
i
al
n
e
u
ral
n
e
t
w
o
r
k
,
"
In
t
J
A
d
v
Co
m
p
u
t
R
e
s
.
,
V
o
l
.
8
,
N
o
.
3
9
,
p
p
.
3
2
4
-
34
,
2
0
1
8
.
[2
6
]
A
rb
o
l
e
d
a
E
.
R
.
,
Faj
ard
o
A
.
C
.
,
Med
i
n
a
R
.
P.
,
"
Cl
as
s
i
fi
cat
i
o
n
o
f
Co
ffee
Bea
n
Sp
ec
i
es
U
s
i
n
g
Imag
e
Pro
ce
s
s
i
n
g
,
A
rt
i
fi
c
i
al
N
e
u
ral
N
et
w
o
r
k
an
d
K
N
eare
s
t
N
e
i
g
h
b
o
rs
,"
2
0
1
8
IE
E
E
I
n
t
Co
n
f
I
n
n
o
v
R
es
D
ev.
,
p
p
.
1
-
5
,
May
2
0
1
8
.
[2
7
]
A
rb
o
l
e
d
a
E
.
R
.
,
Faj
ard
o
A
.
C
.
,
Med
i
n
a
R
.
P.
,
"
A
n
Imag
e
Pro
ces
s
i
n
g
T
ec
h
n
i
q
u
e
fo
r
Co
ffee
Bl
ack
Bean
s
Id
en
t
i
f
i
cat
i
o
n
,"
In
:
2
0
1
8
I
E
E
E
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
er
e
n
ce
o
n
In
n
o
v
a
t
i
ve
R
e
s
ea
r
ch
a
n
d
D
evel
o
p
m
en
t
(ICIR
D
).
IE
E
E
;
p
p
.
1
-
5
,
2
0
1
8
.
[2
8
]
O
l
i
v
e
i
ra
E
.
M
.
D
e,
L
eme
D
.
S
.
,
G
ro
en
n
er
B
.
H
.
,
Ro
d
art
e
M
.
P
.
,
G
u
al
b
ert
o
R
.
,
A
l
v
are
n
g
a
F.
,
"
A
co
m
p
u
t
er
V
ISI
O
N
s
y
s
t
em
fo
r
c
o
ffee
b
e
an
s
cl
a
s
s
i
fi
ca
t
i
o
n
b
as
e
d
o
n
c
o
mp
u
t
a
t
i
o
n
al
i
n
t
e
l
l
i
g
e
n
ce
t
ec
h
n
i
q
u
es
,"
J
F
o
o
d
E
n
g
.
,
2
0
1
5
;
[2
9
]
Fari
d
a
h
,
G
ea
O
.
F.
Pari
k
es
i
t
F.
,
"
Co
ffee
Bea
n
G
rad
e
D
et
erm
i
n
a
t
i
o
n
Ba
s
ed
o
n
Imag
e
Parame
t
er,
"
TE
LKO
M
NIKA
Tel
eco
m
m
u
n
i
ca
t
i
o
n
Co
m
p
u
t
i
n
g
E
l
ect
r
o
n
i
c
s
a
n
d
Co
n
t
r
o
l
,
V
o
l
.
9
,
N
o
.
3
,
p
p
.
5
4
7
-
54
,
2
0
1
1
.
[3
0
]
N
ag
a
s
an
k
ar
T
.
,
A
n
k
ary
ark
a
n
n
i
B.
,
"
Perfo
rman
ce
A
n
a
l
y
s
i
s
o
f
E
d
g
e
D
et
ec
t
i
o
n
A
l
g
o
ri
t
h
m
s
o
n
V
ar
i
o
u
s
Imag
e
T
y
p
es
,
"
I
n
d
i
a
n
J
S
ci
e
m
ce
Tech
n
o
l
.
,
V
o
l
.
9
,
J
u
n
e
2
0
1
6
.
[3
1
]
Su
j
at
h
a
P
.
,
Su
d
h
a
K
.
K.
,
"
Perfo
rman
ce
A
n
a
l
y
s
i
s
o
f
D
i
ff
eren
t
E
d
g
e
D
et
ec
t
i
o
n
T
ec
h
n
i
q
u
es
f
o
r
Imag
e
Seg
me
n
t
a
t
i
o
n
,"
In
d
i
a
n
J
S
ci
Tech
n
o
l
,
.
8
,
J
u
l
y
2
0
1
5
.
[3
2
]
G
u
p
t
a
M
.
,
T
azi
S
.
N
.
,
J
ai
n
A
.
,
"
E
d
g
e
D
et
ec
t
i
o
n
U
s
i
n
g
Mo
d
i
f
i
e
d
Fi
refl
y
A
l
g
o
ri
t
h
m
,"
In
:
2
0
1
4
S
i
x
t
h
In
t
er
n
a
t
i
o
n
a
l
Co
n
f
er
e
n
ce
o
n
C
o
m
p
u
t
a
t
i
o
n
a
l
In
t
el
l
i
g
en
ce
a
n
d
Co
m
m
u
n
i
ca
t
i
o
n
Ne
t
w
o
r
k
s
E
d
g
e
,"
2
0
1
4
.
[3
3
]
El
-
s
ay
e
d
M
.
A
.
,
Sen
n
ari
H
.
A
.
M.
,
"
Mu
l
t
i
-
T
h
res
h
o
l
d
A
l
g
o
r
i
t
h
m
Bas
ed
o
n
H
av
r
d
a
an
d
Ch
ar
v
at
E
n
t
ro
p
y
fo
r
E
d
g
e
D
et
ec
t
i
o
n
i
n
Sat
e
l
l
i
t
e
G
ray
s
ca
l
e
Imag
e
s
,
"
J
S
o
f
t
w
E
n
g
A
p
p
l
.
,
p
p
.
4
2
-
52
,
J
a
n
u
ar
y
2
0
1
4
.
[3
4
]
El
-
s
ay
e
d
M
.
A
.
,
K
h
fa
g
y
M
.
A.
,
"
U
s
i
n
g
Ren
y
i
’
s
E
n
t
r
o
p
y
f
o
r
E
d
g
e
D
et
ec
t
i
o
n
i
n
L
ev
e
l
Imag
e
s
,"
In
t
J
In
t
el
l
Co
m
p
u
t
In
f
.
S
ci
.
,
V
o
l
.
2
,
J
u
l
y
2
0
1
1
.
[3
5
]
G
o
y
al
A
.
,
L
i
Z
.
,
K
i
mm
H
.
,
"
Co
mp
arat
i
v
e
St
u
d
y
o
n
E
d
g
e
D
e
t
ect
i
o
n
A
l
g
o
r
i
t
h
ms
u
s
i
n
g
O
p
en
A
CC
a
n
d
O
p
en
MPI
o
n
M
u
l
t
i
c
o
re
Sy
s
t
em
s
,
"
In
:
2
0
1
7
IE
E
E
1
1
t
h
In
t
er
n
a
t
i
o
n
a
l
S
ym
p
o
s
i
u
m
o
n
E
m
b
e
d
d
e
d
M
u
l
t
i
c
o
r
e
/
M
a
n
y
-
co
r
e
S
ys
t
em
s
-
on
-
Ch
i
p
Co
m
p
a
r
a
t
i
ve
,"
2
0
1
7
.
[3
6
]
Sp
o
n
t
o
n
H
.
,
Card
e
l
i
n
o
J
.
,
"
A
Rev
i
ew
o
f
Cl
a
s
s
i
c
E
d
g
e
D
e
t
ect
o
rs
A
l
g
o
r
i
t
h
ms
Ba
s
ed
o
n
t
h
e
Fi
rs
t
D
er
i
v
a
t
i
v
e,
"
Imag
e
Pro
ces
s
L
i
n
e,
V
o
l
.
5
,
p
p
.
90
-
1
2
3
,
2
0
1
5
.
[3
7
]
Ray
M
.
K
.
,
Mi
t
ra
D
.
,
Sah
a
S.
,
"
Si
m
p
l
i
fi
e
d
N
o
v
e
l
Met
h
o
d
fo
r
E
d
g
e
D
e
t
ect
i
o
n
i
n
D
i
g
i
t
al
Ima
g
es
,
"
In
:
P
r
o
ce
ed
i
n
g
s
o
f
2
0
1
1
I
n
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
ce
o
n
S
i
g
n
a
l
P
r
o
ces
s
i
n
g
,
Co
m
m
u
n
i
c
a
t
i
o
n
,
Co
m
p
u
t
i
n
g
a
n
d
Ne
t
wo
r
ki
n
g
Te
ch
n
o
l
o
g
i
e
s
(ICS
CCN
2
0
1
1
)
,
p
p
.
1
9
7
-
2
0
2
,
2
0
1
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.