Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
10
,
No.
4
,
A
ugus
t
2020
,
pp.
3568
~
35
75
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
10
i
4
.
pp3568
-
35
75
3568
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
Bacteri
al fo
ragin
g optimi
zation b
ased ad
ap
tive
n
eu
ro
fu
zz
y
inferenc
e system
C.
Arul
Mur
u
gan
1
,
G.
Sures
hkum
ar
2
,
Nit
hiyana
nt
h
an
Ka
n
nan
3
,
S
un
il
Tho
ma
s
4
1
Depa
rtment of
El
e
ct
roni
cs
and
Te
l
ec
om
m
unic
ation
Engg
,
Karp
a
gam Col
le
g
e
of
Engi
ne
eri
ng,
Ind
ia
2
Depa
rtment of
Ele
ct
roni
cs
and
I
nstrum
ent
at
ion
Engi
ne
eri
ng
,
Ka
rpa
gam Col
l
ege
of
Engi
n
ee
rin
g,
I
ndia
3
Depa
rtment of
El
e
ct
ri
ca
l
Eng
in
ee
ring
,
Fa
cul
t
y
o
f
Engi
n
ee
ring
,
Rabi
gh,
King
Abd
ula
z
iz
Univ
ersity
,
Saudi
Arabi
a
4
Depa
rtment of
El
e
ct
ri
ca
l
and
E
l
ec
tron
ic
s E
ng
ineeri
ng,
Birla
Inst
i
tut
e
of Tec
hno
lo
g
y
and
Sc
ie
n
ce,
Unite
d
Arab
Em
ira
t
es
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
A
ug
19, 201
9
Re
vised Dec
25, 2
019
Accepte
d
Ja
n 1
1,
2020
Li
fe
of
hum
an
bei
ng
and
ani
m
al
s
depe
nd
on
the
envi
ronm
ent
which
is
surrounded
b
y
pla
nts.
L
ike
hu
m
an
bei
ngs,
plants
al
so
suffer
from
lot
of
disea
ses.
Pl
ant
get
s
aff
ec
t
ed
b
y
complet
e
l
y
in
cluding
leaf,
st
em,
root
,
frui
t
and
flowe
r;
thi
s
aff
ects
the
norm
al
gro
wth
of
the
pla
nt
.
Manua
l
id
ent
ifica
ti
on
and
dia
gnosis
of
pla
nt
d
isea
ses
is
ver
y
diff
ic
ul
t.
T
his
m
et
hod
is
co
stl
y
as
well
as
ti
m
e
-
consum
ing
so
it
is
ine
ffic
ie
n
t
to
be
highly
spec
if
ic.
Plant
pat
holog
y
dea
ls wi
th
th
e
pr
ogre
ss
in
dev
el
o
ping
class
ifi
c
at
io
n
of
p
la
n
t
dise
as
es
and
th
ei
r
ide
nti
f
icati
on
.
T
his
work
cl
ari
fi
es
the
ide
nti
f
ica
ti
on
of
pla
nt
dis
ea
ses
using
le
af
images
caus
ed
b
y
ba
ct
er
i
a,
v
iruses
and
fungus.
B
y
th
is
m
et
hod
it
ca
n
be
ide
n
ti
fi
ed
and
con
trol
the
dise
ase
s.
To
ide
n
ti
f
y
th
e
pla
nt
le
a
f
disea
se
Adapt
iv
e
Neuro
Fuzz
y
Infe
ren
c
e
S
y
st
em
(AN
FI
S)
w
as
proposed.
The
proposed
m
et
hod
show
s m
ore
re
fine
d
resul
ts t
han
th
e
ex
isti
n
g
works
.
Ke
yw
or
d
s
:
Ba
ct
eria fora
gin
g al
gorithm
Feat
ur
e
ex
tr
act
ion
Im
age seg
m
entat
ion
Plant disease
So
ft
co
m
pu
ti
ng
Copyright
©
202
0
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Nithiy
anan
t
ha
n Kanna
n,
Dep
a
rtm
ent o
f El
ect
rical
En
gi
neer
i
ng,
Faculty
of E
ngineerin
g,
King
Abd
ul Azi
z U
ni
ver
sit
y,
Ra
bigh, Sau
di
Ar
a
bia
.
Em
a
il
:
n
m
ajak
nap@
kau.ed
u.sa
1.
INTROD
U
CTION
Indian
Eco
no
m
y
is
necessit
ou
s
in
t
he
ro
l
e
of
ag
ricult
ure.
India
is
a
n
a
gr
ic
ultur
e
base
d
c
ountry
,
wh
e
re
m
or
e
t
ha
n
50%
of
po
pula
ti
on
de
pe
nds
on
a
gri
cultur
e.
Like
hum
an
bein
g
plants
al
so
get
a
ff
ect
e
d
from
diseases,
to
pr
e
ven
t
t
his
fa
rm
e
rs
m
us
t
us
e
s
ui
ta
ble
pestic
ide
s
f
or
c
r
op
s
.
B
ot
h
the
quantit
y
and
qual
it
y
of
crops
get
af
fected
du
e
to
diseases.
Visu
al
o
bse
r
va
ti
on
of
a
pla
nt
is
done
by
plant
pat
ho
l
og
y.
Com
pu
te
r
de
vi
ces
act
as
a
co
ns
eq
ue
ntial
ro
le
in
num
ero
us
a
pp
li
cat
ion
s
s
uch
a
s
def
e
ns
e
,
ag
r
ic
ultur
e,
m
edical
and
e
ngin
eerin
g.
Huge
nu
m
ber
of
resea
rch
has
bee
n
re
ported
[1
-
35]
.
In
e
arl
y
days,
the
e
xperti
se
p
ers
on
in
the
fiel
d
m
a
nu
al
ly
analy
ses
an
d
di
agnosis
the
plant
disease
whic
h
are
to
o
di
f
ficult
to
be
m
on
it
ore
d
an
d
a
lso
it
require
d
m
or
e
tim
e. Plant ser
ve
as a b
ack
bone
in all
asp
ect
s o
f
li
fe,
plants
also su
ff
e
r
fro
m
d
ise
ases wh
ic
h
aff
ect
s the nor
m
al
grow
t
h
of the
pl
ant.
In
this
arti
cl
e,
it
has
been
pr
e
sented
a
n
aut
om
at
ic
so
ft
co
m
pu
ti
ng
a
ppro
ac
h
nam
ed
ANFI
S
to
identif
y
and
cl
assify
di
seases
of
plant
le
aves.
Im
age
segm
entat
ion
is
do
ne
to
sp
li
t
the
le
af
i
m
ag
es
into
sm
a
ll
er
par
ts
for
cl
ear
i
den
ti
ficat
ion
of
a
ff
e
ct
ed
pa
rts
of
le
af.
Im
age
se
gm
entat
ion
is
done
with
t
he
te
xture,
c
olor
an
d
s
hap
e
of
a
n
i
m
age.
Segm
entat
ion
is
the
pr
e
process
ing
te
ch
nique
wh
ic
h
com
pr
ise
s
of
soft
com
pu
ti
ng
an
d
tra
di
ti
on
al
m
et
ho
d.
S
of
t
c
om
pu
ti
ng
is
a
n
a
uto
m
at
ic
m
et
hod
wh
ic
h
is
us
e
d
for
c
omp
le
x
ta
s
k
(i.e.
i
m
age
segm
entat
ion
)
without
hu
m
an
knowle
dge.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Bacteri
al
f
or
ag
ing
opti
miza
ti
on
ba
s
ed
ad
ap
ti
ve ne
uro fuz
zy
infere
nce sy
ste
m
(
C. Ar
ul M
uruga
n
)
3569
2.
RELATE
D
W
ORK
In
t
his
te
ch
nique
[
36]
plant
diseases
sym
pt
om
s
was
ident
ifie
d
an
d
it
w
as
te
ste
d
us
i
ng
al
gorithm
to
achieve
bett
er
accuracy.
C
onduct
an
inv
e
sti
gation
[
37
]
in
cl
assifi
cat
ion
m
et
ho
ds
to
ge
t
rid
of
er
rors
and
al
so
discusse
d
the
m
od
el
s
to
extract
infor
m
at
ion
in
infe
ct
ed
le
af.
In
[
38
]
pro
posed
a
m
et
ho
do
l
og
y
in
fo
ur
diff
e
re
nt
ph
ase
s
to
su
pport
accurate
dif
fer
e
ntiat
ion
of
le
af
diseases
and
cl
assify
the
detect
ion
.
Com
par
is
on
was
m
ade
f
or
se
gm
entat
ion
s
us
in
g
a
uto
m
ated
t
oo
l
a
nd
le
af
sym
pto
m
s
wer
e
ide
ntifie
d
[39].
A
n
in
ven
ti
ve
te
chn
iq
ue
el
ic
it
ed
from
OCM
[40]
was
intr
oduce
d
to
a
naly
ze
the
le
af
m
e
chan
ism
and
the
le
af
spots
pro
vide
sta
bili
ty
in
the
fr
am
ewo
r
k
of
hybri
d
al
gorith
m
s.
In
this
w
ork
[41
]
a
new
a
ppr
oach
was
c
on
si
der
e
d
t
o
e
m
ula
te
the
res
ults
and
the
portio
ns
of
le
af
are
com
bin
e
d
to
ad
dre
ss
the
i
m
age
analy
sis.
It
is
widely
us
ed
to
adap
t
the
se
gm
entat
i
on
in
a
ge
ner
al
m
easur
e
of
bi
as.
T
he
im
ages
we
re
autom
at
i
cal
ly
c
la
ssifie
d
[4
2]
in
a
trai
nin
g
set
to
facil
it
at
e
ac
cur
acy
in
deep
le
arn
in
g
m
od
el
s.
App
ro
ac
hes
involve
d
ha
ve
the
pote
ntial
to
ide
ntify
the
sp
eci
e
s
of
pe
rsp
ect
i
ve
m
od
el
s.
Evalu
at
ion
was
m
ade
for
run
ning
t
i
m
e
and
acc
uracy
al
so
guide
li
nes
we
re
give
n
t
o
ov
e
rc
om
e
the
pro
blem
s
in
fut
ur
e
re
se
arc
h
works
[
43
]
.
In
[44],
bacte
rial
sp
ot
was
asse
s
sed
a
nd
al
s
o
e
valuated
the
seve
rity
of
diseases
wh
ic
h
are
i
nacc
ur
a
te
causin
g
le
sion
s
on
le
a
ves.
Using
s
om
e
m
od
ific
at
ion
s
visu
a
l
m
et
ho
ds
qua
ntify
the
pote
ntial
risk
of
i
nf
e
ct
ion
.
The
diseased
a
reas
of
each
l
eaf
are
colle
ct
ed
to
fa
ci
li
ta
te
dig
it
al
scan
ni
ng.
T
he
a
uthor
pro
pose
d
a
n
al
go
rithm
[4
5]
t
o
ide
ntif
y
the
res
ults
of
m
ai
ze
disease
in
the
detect
ion
of
plant
le
af.
A
data
point
i
n
sam
e
cl
ass
i
s
a
pr
im
e
te
ch
nique
to
en
ha
nce
col
our
ana
ly
sis
in
favor
of
feat
ure
extracti
on.
I
n
this
researc
h
wor
k
[
46
,
47
]
de
picts
a
prot
otype
cal
le
d
e
AG
R
OBOT
to
detect
the cr
op
diseas
e.
3.
RES
EA
R
CH MET
HO
D
In
this
pa
per
the
pla
nt
le
af
disease
is
a
na
ly
zed
an
d
pr
e
dicte
d
by
us
in
g
Ad
a
ptive
N
euro
Fu
zz
y
Infer
e
nce
Syst
e
m
.
Ba
sic
al
ly
hu
m
an
bei
ng
s
cal
culat
e
the
c
har
act
erist
ic
s
of
a
pa
rtic
ular
thin
g
by
c
ompari
ng
the
predeterm
ined
value
obta
ined
by
oth
e
r
a
uthors.
Sim
i
la
rly
syst
e
m
analy
ses
the
sam
e
pr
e
determ
ined
values
to
ob
ta
in
t
he
exp
ect
e
d
res
ul
t.
Feat
ur
e
ext
r
act
ion
is
relat
ed
to
dim
ens
ion
al
it
y
reducti
on.
W
it
h
the
help
of
ANFIS, the
pr
opos
e
d
al
gorithm
achieves high
e
r
acc
ur
a
cy
.
4.
PLANT
DI
SEASE
Si
m
il
ar
to
hum
an
bein
gs
an
d
anim
a
ls
rev
eal
sy
m
pto
m
s
of
s
uffer
f
ro
m
dise
ases.
T
he
e
ntir
e
pla
nt
get
s
aff
ect
ed
by
t
he
diseases
i
nclu
ding
fl
ow
e
r,
r
oot,
le
af,
ste
m
a
nd
f
ru
it
.
M
os
t
of
t
he
ti
m
e
i
t
i
s
dif
ficult
to
i
de
ntify
the
pla
nt
un
de
r
va
rio
us
facto
rs
wh
ic
h
m
ay
cause
le
a
ves
dro
p,
flo
wer
an
d
f
r
uit.
W
el
l
-
de
fine
d
treat
m
e
nt
an
d
identific
at
ion o
f
pla
nt d
ise
ase
s
is require
d for
g
r
owth
of cert
ai
n
fact
or
s a
nd
app
r
opriat
e d
ia
gnos
is.
Th
e
as
pects
for dia
gnosi
s of
diseases
dep
e
nd
s
on
Glancin
g
f
or
s
ym
pto
m
s
or
sign
s:
the
a
ppear
ance
of
so
m
e
un
wa
nted
s
pots,
cur
ls
an
d
dead
areas
are
visi
bl
e
to the na
ked ey
es.
Aw
a
re
ness
of
the
norm
al
char
act
erist
ic
s
of
ho
st
plant:
it
is
eas
ie
r
f
or
one
to
diag
nosis
the
plant
disea
se
if one s
houl
d k
now
t
he
in
he
rita
nces a
nd asset
s of the
host
pl
ants.
The
occ
urre
nc
es
of
sym
pto
m
s
:
i
t
dep
en
ds
upon
two
f
act
or
s
(i
)
the
Diso
r
de
r,
(ii)
Diseases
som
e
env
i
ronm
ental
pro
blem
s
le
ad
to
disorde
rs,
it
happe
n
s
sud
de
nly
li
ke
within
a
day
or
wee
k
and
does
s
pr
e
ad
ov
e
r
t
he parts
of the
plants.
Detect
ing
t
he
diseases
of
t
he
host
pla
nt
in
natur
e
patte
rn
is
of
tw
o
t
ypes
unif
or
m
and
no
n
-
un
i
form
.
Non
-
li
ving
fac
tors
are
t
he
ca
us
e
f
or
unifo
r
m
patte
rn
w
he
reas
no
n
-
un
i
form
patte
rn
is
cause
d
by
som
e
disease
or
i
nse
ct
.
The
disea
se
cl
assifi
cat
ion
an
d
it
s
ef
fec
t,
sy
m
pto
m
s
of
f
ungal
plant
are
ex
plaine
d
in
Table
1.
Table
1
.
Fun
ga
l plant
disease
cl
assifi
cat
ion
Leaf I
m
ag
e
Diseas
es n
a
m
e
Ef
f
ect
Su
itab
le Cli
m
atic
Co
n
d
itio
n
Sy
m
p
to
m
s
/Cau
se
Co
m
m
o
n
r
u
st
Ro
ses
,
h
o
lly
h
o
ck
s,
sn
ap
d
rago
n
s g
et
af
f
ected
Low te
m
p
e
rature
Ap
p
ears pri
m
a
ril
y
o
n
the
su
rf
ace
o
f
lower le
af
,
redd
ish
o
rang
e sp
o
re
m
ass
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3568
-
3575
3570
Table
1
.
Fun
ga
l plant
disease
cl
assifi
cat
ion
(
Con
ti
nue)
Leaf I
m
ag
e
Diseas
es n
a
m
e
Ef
f
ect
Su
itab
le
Cli
m
atic
Co
n
d
itio
n
Sy
m
p
to
m
s
/Cau
se
Late
b
lig
h
t
en
tire
p
lan
ts g
et
in
f
ected
te
m
p
eratur
e belo
w
7
0
d
eg
ree
Fah
renh
eit
rese
m
b
le
on
the lo
wer, o
ld
er
leav
es as wat
er
so
a
k
ed
,
g
rey
g
reen sp
o
ts
Ced
ar
ap
p
le r
u
st
in
f
ect jun
ip
ers
sp
o
re
o
v
er
win
ter
ap
p
ears a
s a
redd
ish
b
rown
gall
m
o
istu
ri
zed weath
er
Leaf cu
rl
redu
ce fru
it
p
rod
u
ctio
n
te
m
p
eratur
e abo
v
e
7
0
d
eg
ree
Fah
renh
eit
leaf
sh
ap
e chan
g
es
in to
sp
ring
,
redd
ish
ar
e
a on
d
ev
elo
p
in
g
need
s
Leaf sp
o
t
as sp
o
ts b
eco
m
e
m
o
re
n
u
m
e
rou
s,
en
tire
leav
es
m
a
y
y
ello
w and
leaf
dr
o
p
will takes
place
war
m
and
coo
l
te
m
p
eratur
e
Leaf sp
o
t,
b
rown
blo
tch
es o
n
th
e leaf
Ear
l
y
blig
h
t
ste
m
,
f
ruit
an
d
up
p
er
p
o
sitio
n
o
f
the p
la
n
t
g
et inf
ected
m
o
d
e
rate
te
m
p
eratur
e
s
m
all bro
wn
sp
o
t a
p
p
ears on
th
e lower side o
f
t
h
e leaves
with
co
n
cent
ric
rin
g
s wh
ich
f
o
r
m
s
a bu
ll
-
s e
y
e
p
attern
5.
PROP
OSE
D WOR
K
In
t
he
pro
po
se
d
work,
it
has
been
sp
i
rite
d
on
ide
ntific
at
ion
an
d
cl
assifi
cat
ion
of
pla
nt
dis
eases
us
i
ng
a
few
c
om
pu
ta
ti
on
al
an
d
intel
li
gen
ce
ad
va
nc
e.
The
perfect
ion
researc
h
a
re
a
is
e
m
erg
ing
f
or
the
de
velo
pm
ent
of
a
n
a
uto
m
ated
syst
em
fo
r
identify
in
g
a
nd
cl
assi
fyi
ng
the
ad
ulterat
ed
pla
nts
wit
h
di
ff
e
ren
t
di
seases.
To
pr
e
ve
nt
the
qual
it
at
ive
an
d
qu
a
ntit
at
ive
loss
of
ag
ricul
tural
yi
el
ds
i
de
ntific
at
ion
of
diseases
is
th
e
only
way.
T
he
lim
itati
on
of
hu
m
an
visi
on
ca
n
be
ov
e
rc
om
e
by
the
accu
rate
and
ti
m
ely
detect
i
on
of
diseas
es
with
the
hel
p
of
im
age
processi
ng
te
chn
i
qu
e
s.
T
he
le
af
im
age
is
pr
e
-
processe
d
init
ia
ll
y.
Usi
ng
ada
ptive
k
-
m
eans
diseases
par
t
is
segm
ented.
T
extu
re
a
nd
sta
ti
sti
cal
te
chn
ique
are
al
s
o
incl
ud
e
d
t
o
ob
ta
in
high
pass
versi
on
of
featur
e
ex
t
racti
on
.
5.1.
Ima
ge acquisi
tion
The
ac
quire
d
i
m
age
is
e
nh
a
nc
ed
by
us
in
g
UN
s
harp
filt
er,
by
filt
ering
a
U
N
s
ha
rp
ve
rsion
i
n
dark
i
m
age
from
the
or
i
gin
al
im
a
ge.
It
filt
ers
t
he
ed
ges.
When
com
par
ed
with
neg
at
ive
of
t
he
La
placi
an
filt
er
the
par
am
et
er
al
pha
is
us
e
d
f
or
unsha
r
p
filt
er
a
nd
t
he
al
pha
va
lue
is
sel
ect
ed
as
0.2,
by
th
e
‘f
s
pecial
’.
K
-
m
eans
m
et
ho
d
is
u
se
d f
or
im
age segm
entat
ion
a
nd
flo
wch
a
rt for
the
pro
po
se
d
a
ppr
oac
h
is s
how
n belo
w
in
Fig
ur
e
1.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Bacteri
al
f
or
ag
ing
opti
miza
ti
on
ba
s
ed
ad
ap
ti
ve ne
uro fuz
zy
infere
nce sy
ste
m
(
C. Ar
ul M
uruga
n
)
3571
Figure
1. Flo
w
char
t
for pr
opose
d
a
ppr
oach
6.
RESU
LT
S
AND
A
N
ALYSIS
The
pro
pose
d
work
was
im
pl
e
m
ented
on
M
ATL
AB
2014a
.
Fo
r
j
ust
ify
ing
the
su
ccess
of
this
wo
r
k
we
ha
ve
ta
ke
n
two
set
of
im
a
ges.
Fi
rst
we
ha
ve
ta
ke
n
6
set
of
im
ages
as
an
in
pu
t
im
ages
wh
ic
h
are
a
ffec
te
d
by
diseases
ca
us
e
d
by
f
ungi
and
s
eco
nd
ste
p
we
ha
ve
fi
na
ll
y
two
hundre
d
an
d
se
ve
nty
i
m
ages.
The
a
f
fected
reg
i
on
of
the
l
eaf
pa
rt
is
iden
ti
fied
by
segm
entat
ion
of
pixe
ls.
The
pr
opose
d
w
ork
is
do
ne
su
cce
ssf
ully
with
the
accu
rate
va
lue
that
is
the
diseases
a
re
c
la
ssifie
d
pe
r
fe
ct
ly
.
Fo
r
cl
assi
ficat
ion
t
he
pa
ram
et
ers
are
re
qu
i
re
d
are
sp
eci
fici
ty
and
Sens
it
ivit
y.
The
af
fected
area
of
le
af
pa
r
t
are
ob
se
r
ved
and
pr
e
sente
d
in
Table
2
with
their
cl
assifi
cat
ion
s.
Table
2.
Res
ults f
or
A
NFIS
,
(
a)
Or
igi
nal im
a
ge
,
(
b)
Segm
ented
gray
scal
e i
m
age
,
(c
)
Ma
s
k
im
age
,
(d)
Re
s
ults f
or
segm
ented
le
af
d
ise
ase
(a)
(b)
(c)
(d)
Co
m
m
o
n
r
u
st
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3568
-
3575
3572
Table
2.
Res
ults f
or
A
NFIS
,
(
a)
Or
igi
nal im
a
ge
,
(
b)
Segm
ented
gray
scal
e i
m
age
,
(c
)
Ma
s
k
im
age
,
(d)
Re
s
ults f
or
segm
ented
le
af
d
ise
ase
(
Co
ntinu
e)
(a)
(b)
(c)
(d)
Late
b
lig
h
t
Ced
ar
ap
p
le r
u
st
Leaf cu
rl
Leaf sp
o
t
Ear
l
y
blig
h
t
Plant
le
af
dise
ases
dep
e
nds
on
it
s
own
f
ea
ture
re
fer
s
to
s
ym
pto
m
s.
Feat
ur
es
li
ke
s
hape,
colo
r
an
d
siz
e
can
be
de
te
rm
ined
by
featur
e
e
xtract
ion
.
To
sea
rc
h
seed
points
a
nd
t
o
gro
up
them
reg
ion
grow
i
ng
m
et
ho
d
is
us
ed
in
featu
re
e
xtr
act
ion
.
F
or
wel
l
-
def
i
ned
e
x
am
inati
on
de
fine
d
the
i
nput
im
a
ge
is
c
onve
rted
int
o
segm
ented
gr
a
y
scal
e
i
m
age.
Sp
eci
fici
ty
and
Sens
it
ivit
y
validat
ion
pa
ra
m
et
ers
us
ed
in
the
pr
op
os
ed
work.
Ad
a
ptive
K
-
m
eans
is
us
e
d
f
or
evaluati
ng
th
e
segm
entat
ion
process.
Ta
bl
e
3
show
s
the
repor
t
of
s
pec
i
fici
ty
for
the
propos
ed
al
gorithm
al
ong
with
t
he
com
par
ison
of
existi
ng
al
go
r
it
h
m
s.
Table
4
sh
ows
the
re
port
of
s
ensiti
vity
fo
r
the
pr
op
os
e
d
al
go
rithm
al
on
g
with
the
com
par
ison
of
existi
ng
al
gori
thm
s.
The
pictorial
represe
ntati
on
of
detect
ed
le
a
f
diseases
for
pro
posed
a
nd
ex
ist
ing
al
gorith
m
s
with
sp
eci
f
ic
it
y
resu
lt
s
is
sh
ow
n
in
Figure
2.
T
he
pictorial
re
pr
ese
ntati
on
of
detect
ed
le
af
diseases
f
or
pr
opos
e
d
an
d
exi
sti
ng
al
gorith
m
s
with
s
ensiti
vity
resul
ts i
s shown i
n Fi
gure
3.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Bacteri
al
f
or
ag
ing
opti
miza
ti
on
ba
s
ed
ad
ap
ti
ve ne
uro fuz
zy
infere
nce sy
ste
m
(
C. Ar
ul M
uruga
n
)
3573
Table
3.
C
om
par
iso
n of speci
fici
ty
r
esults f
or
detect
ion
of l
eaf
diseases
Diseas
e
K
-
m
e
an
s
Gen
etic algo
rith
m
BR
BFNN
ANFIS
Co
m
m
o
n
r
u
st
0
.78
1
7
0
.80
9
6
0
.82
1
3
1
.00
0
0
Late
b
lig
h
t
0
.80
1
4
0
.82
0
5
0
.83
2
6
0
.85
7
1
Ced
ar
ap
p
le r
u
st
0
.78
0
1
0
.78
5
4
0
.81
9
6
1
.00
0
0
Leaf cu
rl
0
.75
1
7
0
.78
8
9
0
.88
7
9
0
.85
7
1
Leaf sp
o
t
0
.81
2
4
0
.83
1
8
0
.88
3
6
1
.00
0
0
Ear
l
y
blig
h
t
0
.82
1
1
0
.83
7
4
0
.88
9
7
0
.85
7
1
Av
erage sp
ecif
icit
y
0
.79
1
4
0
.81
3
9
0
.85
5
8
0
.92
8
5
Table
4.
C
om
par
iso
n of sen
sit
ivit
y resu
lt
s fo
r
d
et
ect
io
n of
le
af
disease
Diseas
e
k
-
m
eans
Gen
etic algo
rith
m
BR
BFNN
ANFIS
Co
m
m
o
n
r
u
st
0
.80
7
8
0
.81
1
7
0
.83
1
1
1
.00
0
0
Late
b
lig
h
t
0
.81
8
9
0
.83
3
9
0
.84
9
7
1
.00
0
0
Ced
ar
ap
p
le r
u
st
0
.78
0
9
0
.79
9
6
0
.84
0
7
0
.85
7
1
Leaf cu
rl
0
.77
2
9
0
.82
7
9
0
.89
3
8
1
.00
0
0
Leaf sp
o
t
0
.81
1
1
0
.83
1
5
0
.90
7
8
1
.00
0
0
Ear
l
y
blig
h
t
0
.82
0
1
0
.84
7
1
0
.89
9
9
1
.00
0
0
Av
erage
sen
sitiv
ity
0
.80
2
0
0
.82
4
4
0
.87
0
5
1
.00
0
0
Figure
2. I
nf
e
r
ence
for
s
pecif
ic
it
y i
nterp
retat
ion
s
Figure
3. I
nf
e
r
ence
for
se
ns
it
ivit
y i
nterpreta
ti
on
s
7.
CONCL
US
I
O
N
If
hum
an
bein
gs
get
aff
ect
e
d
by
diseases,
it
has
been
c
ur
e
d
by
it
sel
f
consulti
ng
a
doct
or,
but
plants
sh
oul
d
be
obse
rv
e
d
by
hum
an
beings
to
prev
ent
the
disease
s.
But
hum
an
ob
s
er
vation
for
a
pr
ol
onge
d
ti
m
e
is
i
m
po
ssible
a
nd
it
is
di
ff
ic
ult.
To
overc
om
e
this,
a
n
aut
om
a
ti
c
appro
a
ch
w
as
de
sig
ned
w
hich
hel
ps
t
o
m
on
it
or
the
plant
le
av
es
and
id
e
ntif
y
the
le
af
disease.
This
auto
m
at
ic
app
r
oac
h
cl
early
identifie
s
the
disease
and
disease
ca
n
be
cur
e
d
at
ea
rlie
r
sta
ge
.
F
or
i
den
ti
fyi
ng
t
he
plant
le
af
dise
ases
A
NFIS
ha
s
bee
n
us
ed
.
Wh
il
e
com
par
ing
to
t
he
existi
ng
al
gorithm
resu
lt
s,
the
pro
po
se
d
a
lgorit
h
m
sh
ows
90
pe
rce
nt
re
su
lt
s.
Pro
po
se
d
wo
r
k
sh
ows
higher
accuracy
an
d
e
ff
ect
ive
ness.
P
rop
os
ed
wor
k
has
bee
n
su
cc
e
ssfu
ll
y
done
on
f
ungal
disea
ses,
in
fu
t
ur
e it
ca
n be
expect
ed
t
o w
ork o
n bacte
ria
and
vir
us
es.
0
1
2
Comm
on
rust
Late
bligh
t
Ced
ar
ap
ple rust
Leaf
curl
Leaf
s
p
ot
Ear
l
y
bligh
t
A
v
erag
e
sensi
tiv
i
ty
Specif
ici
ty
R
esults
k-m
eans
Gen
eti
c
a
l
g
ori
thm
BR
BF
NN
A
NF
IS
0
1
2
Comm
on
rust
Late
bligh
t
Ced
ar
ap
ple rust
Leaf
curl
Leaf
s
p
ot
Ear
l
y
bligh
t
A
v
erag
e
sensi
tiv
i
ty
Sensitivi
ty
R
esults
k-m
eans
Gen
eti
c
a
l
g
ori
thm
BR
BF
NN
A
NF
IS
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3568
-
3575
3574
REFERE
NCE
S
[1]
Sunil
Thomas,
K.Nit
hi
y
an
ant
ha
n,
“
A
Nove
l
m
et
hod
to
implement
MP
PT
Algorit
hm
s
for
PV
pane
ls
on
a
MA
TL
AB/S
IM
ULINK
envi
ronm
ent
,
”
Journal
of
Adv
anc
ed
Re
s
earc
h
in
Dynamic
al
and
Control
Syste
ms
,
vol.
10,
no
4,
pp.
31
-
40
,
2018.
[2]
Gow
rishanka
r.
K.,
Sunil
Thomas,
K.
Nithiy
a
na
ntha
n,
“
W
ire
le
ss
Inte
gr
at
ed
-
Sen
sor
Network
b
ase
d
Subs
ea
Tunn
e
l
Monitori
ng
S
y
s
te
m
,
”
Journal
of
Ad
vanc
ed
R
ese
arch
in
Dyn
amical
and
Co
ntrol
Syst
ems,
vol.
10
,
No
12
,
pp.
647
-
658,
201
8.
[3]
Gow
rishanka
r.
K.
K.Nit
hi
y
an
ant
h
an.
,
M.Pri
y
a
nka
,
Sunil
Thomas,
GS
M
base
d
dua
l
power
enha
nc
e
d
LE
D
display
not
ice
boar
d
wit
h
m
oti
on
detec
t
or,
”
In
te
rnation
al
Journal
o
f
E
ngine
ering
and
Technol
ogy
,
vol
.
7,
pp.
559
-
56
6
,
2018.
[4]
S
y
ahr
e
l
Emran
Bin
Siraj
,
T
an
Yong
Sing,
Raman
Ragura
m
an,
Prata
p
Nair
Marimuthu,
K.
Nithiy
an
ant
ha
n
,
“
Applic
at
ion
of
Cluste
r
Anal
y
s
is
and
As
socia
ti
on
Anal
y
sis
Model
Based
Pow
er
Sy
st
em
Fault
Ide
nti
f
icati
o
n,
”
European
Journ
al
of
Scientifi
c
R
ese
arch
,
vol. 13
8,
pp
.
16
-
28
,
20
16.
[5]
K.
Nithiy
ana
n
th
an,
Prata
p
Nair
,
Raman
Ragura
m
an,
Ta
n
Yong
Sing,
Sy
ah
rel
Emran
Bi
n
Siraj
,
“
Enha
nce
d
R
pac
kag
e
-
base
d
c
luste
r
anal
y
s
is
fau
lt
id
ent
if
ic
a
ti
o
n
m
odel
s
for
th
ree
phase
power
sy
st
em
net
work,”
Inte
rnat
iona
l
Journal
of
Busin
ess Int
ellige
n
ce
and
Data
Min
in
g
,
vol
.
14
,
pp
.
10
6
-
120,
2019
.
[6]
Sekat
h
Varm
a,
K.
Nithiy
ana
n
th
an,
“
MA
TL
AB
Sim
u
la
ti
ons ba
sed
Ide
nti
fi
ca
t
ion
Model
for
Vari
ous Points
in
Globa
l
Pos
it
ioni
ng
S
y
s
t
em,”
In
te
rnation
al
Journal
of
Co
mputer
Applicati
ons
,
vol. 138, no
.
13
,
pp15
-
18
,
2
016.
[7]
S.
Sam
son
Raj
a
,
R.
Sundar,
A.A
m
udha,
K.
Nithi
y
an
ant
h
an,
“
Unit
Com
m
it
m
ent
C
al
cu
la
tor
Mode
l
for
Thre
e
phas
e
Pow
er
Sy
st
em
Network,
”
In
te
r
nati
onal
Journa
l
of
Pow
er
and
Re
newab
le
E
nergy
Syste
ms
,
vol
3,
no
.
1
,
pp.
37
-
42
,
2016
.
[8]
K.
Gow
rishanka
r,
K.
Nithiy
a
nan
tha
n,
Man
i
Pri
y
a
nka,
Venka
te
san
,
G.,
“
Neura
l
n
etw
ork
base
d
m
at
hemati
c
al
m
ode
l
for
fee
d
m
ana
g
e
m
ent
technique
i
n
aquacul
tur
e,”
Journal
of
Adv
a
nce
d
Re
search
i
n
Dynamic
al
an
d
Control
S
ystem
s
,
vol.
18
,
no
.
1,
pp.
1142
-
1161,
201
7.
[9]
Sing,
Ta
n
Yong
and
Bin
Siraj
,
S
y
ahr
el
Emr
an
and
Raguram
an,
Raman
and
Marimuthu,
Prata
p
Nair
and
Nithiy
an
ant
han
,
K.,
“
Cosine
sim
il
arit
y
c
luste
r
an
aly
s
is
m
odel
bas
ed
eff
e
ct
iv
e
po
wer
s
y
stems
fau
lt
ide
n
ti
fi
cation,”
Inte
rnational
Jo
urnal
of Adv
an
c
ed
and
Appl
i
ed
S
ci
en
ce
s
,
vol
4
,
n
o1,
pp
.
123
-
129
,
2017.
[10]
Prata
p
Nair
.
,
Ni
thiy
an
ant
han
,
K.
,
Dhinaka
r
,
P.,
“
Design
and
deve
lopment
of
v
ari
ab
l
e
fre
qu
ency
ul
tra
sonic
p
est
rep
eller
,
”
Journ
al
of
Adv
an
ce
d
Re
search in Dy
n
amical
and
Cont
rol Sy
stems
,
vol. 9, no. 12, pp. 22
-
34,
2017
.
[11]
K.Nit
hi
y
an
ant
ha
n,
Sundar,
R.
,
R
anga
na
tha
n
,
T.,
Sam
son
Raj
a,
S.,
“
Virtua
l
instru
m
ent
at
ion
base
d
sim
ple
ec
onom
ic
loa
d
dispa
tc
h
es
ti
m
at
or
m
odel
f
or
3
-
phase
powe
r
s
y
stem
net
wor
k,
”
Journal
o
f
A
dvanc
ed
Re
sear
ch
in
Dynamic
a
l
and
Control
S
yst
ems
,
vol
.
9
,
no
.
11,
pp
.
9
-
15
,
20
17.
[12]
K.Nit
hi
y
an
ant
ha
n,
T.
Sam
son
Ra
ja
,
Sundar
,
R.
,
Am
udha,
A.,
“
Virtua
l
st
abi
l
ity
esti
m
at
or
m
ode
l
for
three
phas
e
power
s
y
stem
net
work,”
Indon
e
sian
Journal
of
El
e
ct
rica
l
Enginee
ring
and
Co
mputer
Sci
en
ce
,
vol.
4,
no
.
3,
pp.
520
-
525
,
20
16.
[13]
Prata
p
Nair
,
M.
,
K.Nit
hi
y
a
nan
t
han,
“
An
eff
ecti
ve
ca
b
le
siz
ing
proc
edur
e
m
ode
l
for
industries
and
comm
erc
ia
l
buil
dings,
”
Int
ernati
onal Journal
of El
e
ct
ri
cal
an
d
Computer
Eng
ine
ering
,
vol
.
6
,
no.
1
,
pp
.
34
-
39
,
2016.
[14]
Prata
p
Na
ir,
M
.
,
K.Nit
h
i
y
an
anthan,
“
Feasibility
anal
y
sis
m
od
el
for
m
ini
h
y
d
ropower
pla
n
t
i
n
Ti
om
an
Isl
an
d,
Malay
s
ia,”
Distribute
d
Gene
ratio
n
and
A
lt
ernat
ive
En
ergy
Journal
,
vo
l. 31, no. 2,
pp.
36
-
54
,
2016
.
[15]
K.Nit
hi
y
an
ant
ha
n,
Um
asa
nkar
,
“
Envi
ronm
ent
fri
endly
voltage
up
-
gra
dation
m
odel
for
distri
buti
on
power
sy
st
ems
,
”
Inte
rnational
Jo
urnal
of El
e
ct
ri
c
al
and
Comput
er
Engi
n
ee
ring
,
vo
l.
6
,
no
.
6
,
pp
.
25
16
-
2525.
,
2016.
[16]
Sureshkum
aa
r,
G.,
Kann
an,
N
.
,
Thomas,
S.,
Kar
thi
ke
y
a
n,
S.P
.
,
“
Matl
ab/
sim
uli
nk
sim
ula
ti
ons
b
ase
d
m
odifi
ed
sepi
c
DC
to
DC
converte
r
for
ren
ewa
b
l
e
en
erg
y
app
li
c
a
ti
ons,”
Journal
o
f
Adv
an
ce
d
Re
se
arch
in
Dynamic
al
and
Contro
l
Syste
ms
,
vol
.
11
,
no.
4,
pp.
285
-
2
95,
2019
.
[17]
G.
Venka
t
esa
n,
K.
Nithiy
a
nant
h
an,
“
Static and
T
ran
sient
The
rm
al Analy
s
is of
Pow
er
Fact
or
Conv
ert
ers
in
Sw
it
ch
e
d
Rel
uctanc
e
Motor
Drive
,
”
Jour
nal
of
Adv
anced
Re
search
in
Dynamic
al
&
Control
Syste
ms
,
vol
10,
no
13,
pp.
1656
-
1662,
2
018.
[18]
K.Nit
hi
y
an
ant
ha
n,
Ramacha
ndr
a
n,
V.,
“
Eff
ectiv
e
dat
a
compress
io
n
m
odel
for
on
-
li
ne
power
s
y
ste
m
,
”
Inte
rnationa
l
Journal
for
Enginee
ring M
od
el
l
i
ng
,
vol
.
27
,
no
.
3
-
4,
pp
.
101
-
109
,
2014.
[19]
K.Nit
hi
y
an
ant
ha
n,
Rama
cha
ndra
n,
V.
,
“
Distributed
m
obil
e
ag
ent
m
odel
for
m
ulti
ar
ea
power
s
ystems
a
utomate
d
onli
ne
sta
te
estim
at
ion,
”
In
te
rn
ati
onal
Journal
of
Computer
Aided
Engi
n
ee
ring
and
Tech
nolog
y
,
vo
l.
5
,
no.
4
,
pp.
300
-
310
,
20
13.
[20]
K.Nit
hi
y
an
ant
ha
n,
Ramach
andr
a
n,
V.,
“
Versioning
-
base
d
servi
c
e
-
orie
n
te
d
m
ode
l
for
m
ult
i
-
are
a
power
s
y
st
em
onli
ne
ec
onom
i
c
loa
d
dispatch,
”
Computers and Electric
a
l
Eng
in
ee
ring
,
vol
.
39
,
no.
2
,
pp
.
433
-
4
40,
2013
.
[21]
K.Nit
hi
y
an
ant
ha
n,
Ramach
andr
a
n,
V.,
“
Loc
a
ti
on
inde
pend
ent
distri
bute
d
m
odel
f
or
on
-
li
ne
lo
ad
f
low
m
onit
orin
g
for
m
ult
i
-
ar
ea po
wer
s
y
st
ems
,
”
In
te
rnational
Jour
nal
fo
r
Engi
ne
ering
Mode
ll
ing
,
v
ol.
24
,
no
.
1
-
4
,
p
p.
21
-
27
,
2011
.
[22]
K.Nit
hi
y
an
ant
ha
n,
Loomba,
A.
K.,
“
MA
TL
AB/S
imulink
base
d
spee
d
cont
rol
m
odel
for
convert
er
cont
rol
le
d
D
C
drive
s,
In
te
rnat
i
onal
Journal
for
Engi
ne
ering
Mo
del
li
ng
,
no
.
24
n
o.
1
-
4
,
pp
.
49
-
56
,
2011
.
[23]
Thomas,
S.,
K.
Nithiy
an
ant
han
.
,
Eski,
A.
,
“
Inve
s
ti
gations
on
tra
n
sient
beha
v
ior
of
an
ene
rg
y
cons
erv
ation
choppe
r
fed
DC
serie
s
m
otor
subjec
t
ed
to
a
cha
ng
e
in
dut
y
c
y
c
l
e
,”
Journal
of
Gr
ee
n
Engi
n
ee
ring
,
vo
l.
9
,
no.
1
,
pp.
92
-
11
1
.
2019
.
[24]
K.Nit
hi
y
an
ant
ha
n,
Ram
ac
handr
a
n,
V.,
“
A
plug
and
play
m
odel
for
JIN
I
base
d
on
-
li
ne
rela
y
c
ontrol
for
power
s
y
stem prot
ec
t
io
n,
”
In
te
rnat
ional J
ournal
for E
ng
ine
ering
Mode
l
ling
,
vol
.
21
,
no
.
1
-
4
,
pp
.
65
-
68
,
20
08.
[25]
K.Nit
hi
y
an
ant
ha
n,
Ramach
andr
a
n,
V.,
“
A
distri
b
ute
d
m
odel
for
c
apa
c
i
ta
n
ce
r
equirem
ent
s
for
self
-
exc
i
te
d
indu
ct
io
n
gene
ra
tors,
”
Int
e
rnational
Journal
of
Aut
omat
ion and Cont
rol
,
vol
.
2
,
no
.
4
,
pp
.
51
9
-
525,
2008
.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Bacteri
al
f
or
ag
ing
opti
miza
ti
on
ba
s
ed
ad
ap
ti
ve ne
uro fuz
zy
infere
nce sy
ste
m
(
C. Ar
ul M
uruga
n
)
3575
[26]
Nithiy
an
ant
han
.
K,
Ramacha
ndr
an.
V,
“
Com
ponen
t
Model
Sim
ula
ti
ons
for
Mult
i
-
Area
Pow
er
Sy
stem
Model
fo
r
On
-
Li
ne
E
con
o
m
ic
Loa
d
Disp
at
ch
,”
In
te
rnati
onal
Journal
f
or
Eme
rging
E
le
c
tric
Pow
er
Syste
ms
,
vol
1
,
n
o
2,
2004.
[27]
K.
Nithiy
a
nanthan.
,
Elave
n
il
.
V
.
,
“
CYM
GRD
Based
Eff
ective
Ea
r
thl
ing
Design
Model
for
Subs
ta
ti
on
,
”
Inte
rnational
Jo
urnal
for
Compu
te
r A
pp
li
ca
ti
ons
in
Eng
ine
ering
S
ci
en
ce
s
Asia,
v
ol
.
1
,
n
o
3,
p
p.
341
-
346,
2011
.
[28]
K.Ni
thiy
an
ant
ha
n,
Ramacha
ndra
n,
V.,
“
Distribut
ed
m
obil
e
agent
m
odel
for
mul
ti
-
a
rea
power
sy
st
em
on
-
li
n
e
ec
onom
ic
loa
d
d
ispat
ch
,
”
Inte
rna
ti
onal Journal fo
r E
ngineering
M
odel
li
ng
,
vo
l. 17, no. 3
-
4
,
pp.
87
-
90,
2004
.
[29]
Yous
sef
A.
Mo
bar
ak,
A
.
M.
He
m
ei
da,
Nith
i
y
an
ant
han
Kann
an,
Volta
ge
and
Freque
nc
y
b
ase
d
Loa
d
Depe
nd
e
nt
Anal
y
sis
Model
for
Eg
y
ptian
Pow
er
Sy
st
em
N
et
work,
Journal
of
Advanc
ed
Resea
rch
in
D
y
na
m
ic
al
&
Contro
l
S
y
stems
,
US
A. V
ol
11,
No 6, 20
19
Pp.
971
-
978.
[30]
Low
Kak
Hau,
Gow
rishanka
r,
Nithiy
an
ant
han
.
K,
“
Deve
lopment
Of
Protot
y
pe
Model
For
W
ireless
Based
Cont
rol
Pick
And
Place
Roboti
c
Vehi
cl
e
,
”
TEL
KOMNIKA
Indone
si
an
Journal
of
El
e
ct
rica
l
Enginee
ring
,
vo
l.
1
4
,
no.
1
,
pp
110
-
11
5,
April
2015
.
[31]
K.Nit
hi
y
an
ant
ha
n,
Ramach
andr
a
n,
V.,
“
A
distri
b
ute
d
m
odel
for
m
ult
i
-
area
powe
r
s
y
stems
on
-
li
n
e
d
y
namic
se
cur
i
t
y
ana
l
y
sis,
”
I
EE
E
Re
gion
10
Annu
al
Int
ernati
onal
Confe
renc
e, P
ro
ce
ed
ings/TENCON,
vol.
3,
pp.
1
765
-
1767,
2002
.
[32]
K.Nit
hi
y
an
ant
ha
n,
Ramach
andr
a
n,
V.,
“
EJB
base
d
component
m
odel
for
distr
ibut
e
d
loa
d
flow
m
onit
oring
of
m
ulti
-
are
a
power
s
y
st
e
m
s,
”
Inte
rnation
al
Journal
for
E
ngine
ering
Mod
el
li
ng
,
vo
l. 15, n
o.
1
-
4
,
pp
.
63
-
67
,
2002
.
[33]
Sureshkum
aa
r,
G.,
K.Nit
hi
y
an
a
ntha
n.
,
Thomas,
S,
“
M
ATLAB/SIMU
LINK
base
d
sim
ula
ti
ons
of
KY
conve
rte
r
f
or
PV
pane
ls
powe
red
l
ed
li
gh
ti
ng
s
y
stem,
”
Inte
rna
ti
onal
Journal
o
f
Powe
r
El
e
ct
ro
nic
s
and
Dr
iv
e
Syste
ms
,
vol
.
10
,
no.
4
,
pp
.
1885
-
1893,
2019
.
[34]
Vim
al
Raj
,
S.,
S
ure
sh
Kum
ar,
G
.
,
Thomas,
S.,
K.Nit
hi
y
an
ant
ha
n,
“
MA
TL
AB/S
IMU
LINK
base
d
sim
ula
t
ions
o
n
stat
e
of charg
e
o
n
batter
y
for elec
tri
c
al
v
ehicle
s,
”
Journal
of
Gr
een E
ngin
ee
ring
,
v
ol.
9
,
no
.
2
,
pp
.
2
55
-
269,
2019
.
[35]
Mobara
k,
Y.A.,
K.Nit
hi
y
an
ant
ha
n,
“
Vert
ic
all
y
in
te
gra
te
d
utilit
y
p
ower
sy
st
em
struct
ure
s
for
eg
y
p
ti
an
sce
n
ari
o
an
d
el
e
ct
ri
ci
t
y
a
ct
,
”
Journal
of
Adv
anc
ed
R
ese
arch
in
Dyna
mical
and
Co
ntrol
Syste
ms
,
vol.
11,
no
.
7,
pp.
766
-
772
.
20
19
.
[36]
A.
Camargo,
et
al
.
,
“
An
image
proc
essing
b
ase
d
al
gor
it
hm
to
aut
om
at
i
ca
l
l
y
ide
nt
if
y
pla
n
t
dis
ea
se
visu
a
l
s
y
m
ptoms
,
”
Bi
o
systems E
nginee
ring
,
vol
.
102
,
n
o.
1
,
pp
.
9
-
21
,
2
009.
[37]
D.
Sabine
Bauer,
et
al
.
,
“
The
Potent
ial
of
Autom
at
ic
Methods
of
cl
assific
a
ti
o
n
to
ide
nti
f
y
L
e
af
disea
ses
from
Multi
spec
tr
al i
m
age
s”,
Pre
ci
sion
Agric
,
v
ol
12,
pp
.
361
-
377,
2011.
[38]
D.
Al
-
Bashish,
e
t
al
.
,
“
Detect
ion
and
cl
assificat
i
on
of
le
af
dis
ea
s
es
using
K
-
m
ea
ns
-
base
d
segm
ent
at
ion
and
neur
al
net
works
base
d
cl
assifi
ca
t
ion,”
I
nform.
Technol. J
.
,
vol
.
10
,
pp
.
2
67
-
275,
2011
.
[39]
A.
Clé
m
ent,
e
t
al
.
,
“
A
New
Colour
Vision
S
y
st
em
to
Quanti
f
y
Autom
at
ic
a
lly
F
oli
ar
Disco
louration
Caused
b
y
Insec
t
Pests
Feedi
n
g
on
Le
af
Cel
ls,
”
Bi
osystem
s
Engi
nee
ring,
vol.
133,
pp.
1
28
–
140,
2015.
ISS
N
1537
-
5110
,
doi:
10
.
1
016/j.bi
os
y
stems
eng.
20
15.
03.
007
[40]
R.
Zhou,
et
a
l
.
,
"D
isea
se
Detec
ti
on
of
Cer
cospora
Leaf
Spot
i
n
Sugar
Bee
t
b
y
Robust
T
emplate
Matc
h
ing",
Computers and Electroni
cs
in A
gricul
ture
,
vol
.
1
08,
pp
.
58
-
70
,
2
014.
[41]
J.G.A
Barbe
do,
et
al
.
,
“
A
New
Autom
at
ic
Meth
od
for
Disea
se
S
y
m
ptom
Seg
me
ntation
in
Digita
l
Photographs
of
Plant
Le
av
es,
”
E
uropean
Journal
of Pl
ant
Pa
thology
,
vo
l. 147, no
.
2
,
pp
.
349
–
364
,
2016.
[42]
D.
P
Hughes,
et
al
.
,
"A
n
Open
A
cc
ess
Reposit
or
y
of
Im
age
s
on
Plant
Hea
lt
h
to
En
abl
e
the
Deve
lo
pm
ent
of
Mobile
Disea
se
Diagnos
ti
cs
,
”
Corn
el
l
Univer
st
y
2015
.
[43]
A.
Borji,
et
al
.
,
“
Sali
e
nt
Obj
e
ct
De
tecti
on
:
A
Benc
hm
ark
,
”
I
EE
E
Tr
ans.
Image
Proc
ess
.,
v
ol.
24
,
no
.
12
,
pp.
5706
-
5722
,
Dec
.
2015.
[44]
Bardsley
,
et al
.
,
“
Rel
ia
bi
li
t
y
and ac
cur
acy
o
f
visu
al
m
et
hods t
o
qu
ant
if
y
sev
erit
y
o
f
foli
a
r
ba
ct
er
ial
spot
s
y
m
ptom on
pea
ch
and
ne
ct
ar
ine
,
”
Pl
ant
Pat
h
ology
,
vol
.
62
,
n
o.
2
,
2012
.
[45]
Zha
ng,
e
t
al
.,
"O
rthogona
l
Loc
a
lly
Discr
i
m
ina
nt
Projec
tion
for
Cla
ss
ifi
ca
t
ion
of
Plant
Le
af
Disea
ses"
,
Nint
h
Int
ernati
o
nal
Conf
ere
nce
on
Computati
on
al
Int
el
l
ige
nc
e
a
nd
Sec
uri
ty
,
pp
.
241
-
245,
2013.
[46]
Pill
i,
et
al
.
,
"e
AG
ROBO
T
-
A
r
obot
for
e
arly
cro
p
dise
ase
de
te
c
ti
on
using
i
m
age
proc
essin
g,
”
In
te
rnationa
l
Confe
renc
e
on
E
le
c
tronic
s and
Comm
unic
ati
on
S
yste
ms
(
ICECS)
,
pp.
1
-
6,
2014.
[47]
Nithiy
an
ant
han
.
K
,
Mobara
k,
Y.,
Alhar
bi,
F.Applicat
ion
of
cl
oud
computing
for
ec
onom
ic
loa
d
dispat
ch
and
uni
t
comm
it
m
ent
computat
ions
of
the
power
sy
st
e
m
net
work
,
Adv
a
nce
s
in
Inte
llige
nt
Syste
ms
and
Computing
,
1108
AIS
C,
2020
pp.
1
179
-
1189.
Evaluation Warning : The document was created with Spire.PDF for Python.