Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
24
,
No.
2
,
N
ov
em
ber
20
21
,
pp.
10
9
1
~
1099
IS
S
N:
25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v
2
4
.i
2
.
pp
10
9
1
-
109
9
1091
Journ
al h
om
e
page
:
http:
//
ij
eecs.i
aesc
or
e.c
om
IFSG:
I
ntellig
ence ag
ricul
ture crop
-
pest
detecti
on
sy
stem
using
I
o
T
aut
om
atio
n
s
ystem
Imrus S
alehin
1
, S
. M
. Nom
an
2
, Baki
Ul
-
Islam
3
, Isr
at
Jah
an
L
opa
4
, P
ro
dipto Bi
shnu
Ango
n
5
,
Umm
ya
H
ab
i
ba
6
, N
az
mun
Ness
a Moon
7
1,7
Depa
rtment
of
CS
E,
D
aff
odil I
nte
rna
ti
ona
l
Uni
ver
sit
y
,
Dhaka
,
Bangl
ad
esh
2
Depa
rtment of
EE
E
,
Daffod
il In
te
rna
ti
ona
l
Univ
ersity
,
Dhaka
,
B
angl
ad
esh
3
Depa
rtment of
EE
E
,
Pabn
a
Uni
ver
sit
y
of
Sci
ence
and
T
ec
hnolog
y
,
Pabna
,
B
angla
des
h
4
Depa
rtment of
EE
E
,
M
y
m
ensingh
Engi
n
ee
ring
Coll
ege,
M
y
m
en
singh,
Bang
la
d
e
sh
5,6
Facul
t
y
of
Agr
ic
ult
ur
e, Ba
ng
ladesh Agric
ul
tural
Univer
si
t
y
,
M
y
m
ensingh,
Banglade
sh
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
u
n
16
,
2021
Re
vised
A
ug
25
,
2021
Accepte
d
Se
p
30
,
2021
T
h
e
a
g
r
i
c
u
l
t
u
r
a
l
a
n
d
t
e
c
h
n
o
l
o
g
i
c
a
l
c
o
m
b
i
n
a
t
i
o
n
i
s
b
l
e
s
s
e
d
f
o
r
m
o
d
e
r
n
w
o
r
l
d
l
i
f
e
.
I
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s
(
I
o
T
)
i
s
e
s
s
e
n
t
i
a
l
f
o
r
c
o
m
f
o
r
t
a
n
d
d
e
v
e
l
o
p
m
e
n
t
t
o
o
u
r
a
g
r
i
c
u
l
t
u
r
e
s
i
d
e
.
I
n
o
u
r
s
t
u
d
y
,
w
e
d
e
t
e
c
t
e
d
t
h
e
v
a
r
i
o
u
s
p
e
s
t
u
s
i
n
g
d
i
f
f
e
r
e
n
t
t
y
p
e
s
o
f
s
e
n
s
o
r
s
a
n
d
t
h
i
s
i
n
f
o
r
m
a
t
i
o
n
h
a
s
a
u
t
o
m
a
t
i
c
a
l
l
y
s
e
n
t
t
o
t
h
e
f
a
r
m
e
r
'
s
m
o
b
i
l
e
f
o
r
t
h
e
a
l
e
r
t
.
A
l
l
t
h
e
s
e
s
e
n
s
o
r
s
h
a
d
a
c
e
n
t
r
a
l
d
a
t
a
b
a
s
e
.
T
h
o
s
e
s
e
n
s
o
r
s
c
o
l
l
e
c
t
a
l
l
t
h
e
d
a
t
a
a
n
d
d
i
s
p
l
a
y
t
h
e
r
e
s
u
l
t
s
c
o
m
p
a
r
e
d
t
o
t
h
e
c
e
n
t
r
a
l
d
a
t
a
.
T
h
e
H
i
g
h
-
i
m
a
g
e
s
e
n
s
o
r
w
i
l
l
b
e
a
b
l
e
t
o
d
e
t
e
c
t
a
l
l
t
h
e
r
a
y
s
e
m
i
t
t
e
d
f
r
o
m
t
h
e
p
l
a
n
t
a
n
d
a
n
o
t
h
e
r
o
n
e
i
s
t
h
e
g
a
s
s
e
n
s
o
r
w
h
i
c
h
i
s
a
b
l
e
t
o
d
e
t
e
c
t
a
l
l
t
h
e
g
a
s
e
s
c
o
m
i
n
g
f
r
o
m
t
h
e
d
i
s
e
a
s
e
d
p
l
a
n
t
.
W
e
m
a
i
n
l
y
u
s
e
s
o
u
n
d
s
e
n
s
o
r
,
M
Q
1
3
8
,
C
M
O
S
O
V
-
7
6
7
0
,
A
M
G
-
8
8
3
3
f
o
r
a
b
e
t
t
e
r
a
u
t
o
m
a
t
i
o
n
s
y
s
t
e
m
.
W
e
t
e
s
t
i
t
w
i
t
h
r
e
a
l
-
t
i
m
e
e
n
v
i
r
o
n
m
e
n
t
c
o
n
d
i
t
i
o
n
s
(
4
0
°
C
≤
T
A
≤
1
4
°
C
)
.
C
r
o
p
p
e
s
t
d
e
t
e
c
t
i
o
n
a
u
t
o
m
a
t
i
c
p
r
o
c
e
s
s
i
s
m
o
r
e
e
f
f
i
c
i
e
n
t
t
h
a
n
t
h
e
o
t
h
e
r
d
e
t
e
c
t
i
o
n
p
r
o
c
e
s
s
a
c
c
o
r
d
i
n
g
t
o
t
e
s
t
i
n
g
o
u
t
p
u
t
.
A
s
a
r
e
s
u
l
t
,
f
a
r
-
r
e
a
c
h
i
n
g
c
h
a
n
g
e
s
i
n
t
h
e
a
g
r
i
c
u
l
t
u
r
a
l
s
e
c
t
o
r
a
r
e
p
o
s
s
i
b
l
e
.
T
o
r
e
d
u
c
e
e
x
t
r
a
c
o
s
t
a
n
d
i
n
c
r
e
a
s
i
n
g
m
o
r
e
f
a
r
m
i
n
g
a
b
i
l
i
t
y
w
e
n
e
e
d
t
o
I
o
T
a
n
d
A
g
r
i
c
u
l
t
u
r
e
c
o
m
b
i
n
a
t
i
o
n
s
m
o
r
e
.
Ke
yw
or
ds:
Agricult
ure
Detect
ion
syst
em
GS
M m
od
ule
IoT
IS
F
G
se
nsor
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
B
Y
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Im
ru
s Salehi
n
Dep
a
rtm
ent o
f C
SE
Daffodil
Inter
na
ti
on
al
U
ni
ver
s
it
y
Dh
a
ka, B
an
gla
des
h
Em
a
il
: im
ru
s1
5
-
8978@
diu.e
du.bd
1.
INTROD
U
CTION
Nowa
days,
the
t
echn
ol
og
ic
al
revoluti
on
is
a
gr
eat
blessin
g
for
hum
anity.
Si
m
i
la
rly
,
Fo
od
is
ver
y
essenti
al
f
or
hum
an
li
fe
w
hic
h
de
pends
on
a
gr
ic
ultur
e
re
vu
lsi
on
.
F
or
the
great
re
voluti
on
we
ha
ve
pro
pose
d
a
colla
borati
on
betwee
n
a
gri
c
ultur
e
a
nd
i
ntern
et
of
thi
ngs
(
I
oT
)
[1
]
.
T
his
com
bin
at
io
n
is
ver
y
help
fu
l
f
or
far
m
ers
to
culti
vate.
In
ou
r
stud
y,
we
hav
e
de
velo
ped
a
n
au
tom
a
ti
on
crop
pest
detect
i
on
[2
],
[
3]
syst
e
m
based
on
the
im
ag
e
,
so
un
d,
fl
uores
cence,
an
d
ga
s
base
(
I
SFG
)
s
ens
or
m
et
ho
d.
To
ide
ntific
at
ion
plant
disea
ses
or
pest
at
ta
cks,
w
e
dev
el
op
a
n
a
uto
m
at
ion
proc
ess.
This
proce
ss
is
act
ive
when
any
ha
rm
fu
l
causes
are
a
ffec
te
d
by
pests.
IS
F
G
se
nsor
m
et
ho
d
is
a
ne
wly
inv
e
nted
m
et
hod
w
hich
is
we
propose
d.
IS
F
G
re
fer
s
t
o
i
m
agin
g,
so
un
d,
flu
ores
cence,
gas
bas
e
sens
or
inte
grat
ion
syst
em
.
Crop
pests
a
re
anim
a
ls
or
pla
nts
that
dam
age
ob
j
ect
crops
in
the
fa
rm
s.
Trees
are
us
ually
infest
ed
by
insect
s,
bac
te
ria,
an
d
f
ungi.
T
hey
are
a
seriou
s
t
hr
e
at
and
avail
to
over
30
-
50%
retren
c
hm
ent
in
fa
rm
yi
el
d.
F
or
the
detect
io
n
m
es
sage
tra
nsfer
,
we
us
e
global
syst
e
m
for
m
ob
il
e
co
m
m
un
ic
at
ion
(
GS
M
)
m
od
ule
an
d
a
wireles
s
fideli
ty
(
Wi
-
Fi
)
m
on
it
or
in
g
m
od
ule
syst
em
.
Our
ma
in f
oc
us
es
a
re
giv
e
n:
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
2
4
, N
o.
2
,
N
ove
m
ber
20
21
:
10
9
1
-
109
9
1092
Sensor
-
ba
sed
pest
d
et
ect
ion s
yst
e
m
.
Agricult
ural
cr
op h
eal
th
m
on
it
or
in
g base
d o
n
im
age sen
sin
g.
Io
T
b
a
sed
au
t
om
at
ic
m
essage tran
s
fer
syst
e
m
.
GS
M
netw
ork sy
stem
b
ase int
egr
at
io
n
ci
rcu
it
d
e
vice inte
gr
a
ti
on
.
Cost re
duct
ion
and h
i
gh l
onge
vity
d
evice
d
e
ve
lop
f
or
t
he
al
e
rt and
detect
io
n
m
et
ho
d.
This au
tom
at
ion
syst
e
m
is
twop
e
nny and
m
ade w
it
h
ad
va
nc
ed
te
chnol
og
ic
al
m
at
eria
ls. Wireless d
at
a
acce
ss is a w
el
l
-
plan
ne
d
get
-
way to sa
ve
ti
m
e so
that we a
tt
ach th
is m
od
ule [4]
. Th
e
i
ntern
et
of
t
h
in
gs
(IoT)
is
on
e
of
t
he
be
st
te
chnolo
gy
in
this
e
ra
s
o
that
we
c
ou
l
d
be
a
be
ne
fit
to
us
e
it
f
or
e
xcell
ent
li
f
e
an
d
te
chnolo
gical
e
xp
e
diti
on
.
2.
LIT
ERATUR
E REVIE
W
In
sect
s
pose
a
sign
ific
a
nt
risk,
al
so
cause
tw
o
ty
pes
of
ha
r
m
to
cro
ps
in
the
gro
wth
sta
ge
.
The
eat
in
g
insect
s,
w
hich
eat
le
aves
or
bur
rows
in
ste
m
s,
fruit
,
or
r
oots,
cause
direct
dam
age
to
th
e
plant.
The
se
cond
so
rt
of
dam
age
is
ind
irect
dam
age,
in
w
hi
ch
the
insect
c
auses
li
tt
le
or
no
har
m
to
the
crop
but
spr
eads
a
bacteria
l,
vi
ral,
or
fun
gal
il
lness.
Wh
e
n
it
co
m
es
to
rode
nts,
seeds,
le
aves,
roots,
c
om
plete
young
pla
nts,
fruit
,
and
gr
ai
n
are
al
l
on
the
m
en
u.
S
ug
a
r
beet
and
po
ta
t
o
vir
us
es,
for
exam
ple,
are
s
pr
ea
d
by
aph
i
ds
fro
m
on
e
plant
to
the
ne
xt.
Li
et
al
.
i
n
their
resea
rc
h,
im
age
seg
m
entat
ion
al
gori
thm
s
wer
e
appl
ie
d
to
segm
ent
the
destinat
io
n
m
otive
an
d
pr
e
pro
cessi
ng
m
et
ho
d
deals
wit
h
th
e
i
m
ages
that
ha
ve
c
onsidera
bl
e
extent
diff
e
r
ences
betwee
n
the
co
lo
r
of
the
pe
st
and
t
he
co
nd
it
ion
s
.
The
m
ai
n
syst
e
m
s
include
thres
ho
l
ding,
con
t
our
detect
ion,
and
waters
hed
al
go
rithm
[5
]
.
Nag
ar
a
nd
S
ha
rm
a
in
their
research,
the
m
ai
n
obj
ect
ive
of
this
m
et
ho
d
pest’
s
detect
ion
of
plant
pa
rts
s
uc
h
as
r
oo
ts
an
d
le
aves
us
in
g
c
a
pt
ur
in
g
of
pla
nt
le
aves
as
data
colle
ct
ion
an
d
the
n
featur
e
e
xtract
ion
.
Prop
e
r
set
up
of
t
he
wir
e
le
ss
ca
m
era
ne
twork
wh
ic
h
is
connecte
d
w
it
h
Sti
cky
trap
s
for
insect
pests
ca
pturin
g.
CISC
O
Lin
ks
ys
W
i
r
el
ess
-
G
cam
era
was
use
d
as
l
at
est
te
chn
iq
ue
.
The
filt
ering
process
cl
ears
the
nois
e
from
the
i
m
a
ge
ap
pea
ran
ce
du
e
t
o
va
riable
li
gh
ti
ng
c
ondi
ti
on
s
wit
h
im
a
ge
ext
racti
on
m
et
hod
for
the
outp
ut
i
m
age
[6
]
.
D
urga
bai
et
al
.
in
their
resea
rch,
the
yi
el
d
pro
du
ct
io
n
ha
s
co
nd
e
ns
e
d
du
e
to
nu
m
erous
infl
ue
nces
li
ke
pest
at
ta
ck
,
diseases,
and
cl
im
atic
su
rro
undings
.
Crop
protect
io
n
is
the
sci
ence
an
d
rep
et
it
ion
of
s
up
e
r
vision
pes
ts,
plant
disea
ses,
a
nd
oth
e
r
pest
cr
eat
ur
e
s
that
dam
age
agr
ic
ultur
al
c
rop
s
.
Ma
chine
le
a
rn
i
ng
is
a
n
im
m
i
nen
t
fiel
d
of
c
om
pu
te
r
sci
e
nc
e
that
ca
n
be
app
li
e
d
to
the
agr
ic
ultur
al
se
gm
ent
qu
it
e
ef
fecti
ve
ly
.
SetAct
ion
T
hr
es
holds
,
Mo
nitor
a
nd
I
dent
ify
Pests,
Pr
e
ven
ti
on,
C
ontr
ol
are
the
fo
ll
ow
i
ng
ste
ps
.
Bu
g
det
ect
ion
us
i
ng
i
m
ages
of
cr
op
le
aves
has
be
en
em
plo
ye
d
usi
ng
a
patte
r
n
recog
niti
on
br
anch
of
m
achine
le
arn
i
n
g
[7
]
.
Wang
et
al
.
in
their
stud
ie
s,
w
hitefl
ie
s
abd
om
ens
are
ye
ll
ow
an
d
their
wings
are
the
tran
qu
il
sta
ge
t
o
detect
,
m
at
ur
e
a
du
lt
wh
it
ef
ly
was
sel
ect
ed
as
t
he
ta
r
get
insect
at
f
ull
grow
t
h
i
n
this
stud
y.
Veins
a
re
the
va
scular
ti
ssue
of
a
le
af
that
ha
s
a
l
igh
te
r
s
ha
de
,
so
w
hen
se
gm
enting
a
le
af
i
m
age
veins
m
ay
be
detect
ed
as
whit
efly
by
the
al
gorithm
.
Thr
ee
dig
it
al
m
or
phologica
l
featu
r
es
of
a
n
el
li
pse
,
m
ajo
r
an
d
m
ino
r
axis
le
ngth
s,
a
nd
ecce
ntrici
ty
,
are
com
m
on
ly
us
e
d
to
rem
ov
e
veins
f
or
m
i
m
ages.
T
he
m
ai
n
m
i
scal
culation
of
their
researc
h
was
se
gm
entat
ion
occ
urre
d
w
hen
the
w
hitefl
ie
s
or
the
e
gg
s
overla
pp
e
d
wi
th
the
vei
ns
be
cause
the
pro
posed
m
et
ho
d
was
unable
to
deal
eff
ect
ively
with
this
sit
uation
[
8].
Brunel
li
et
al
.
in
their
st
udie
s,
a
n
autom
at
ic
pr
oc
ess
wa
s
occ
up
i
ed
f
or
m
on
it
or
ing
pa
rasit
e
ins
ect
s
from
i
m
ages
ta
ken
i
n
pes
t
traps
as
well
as
an
intel
li
gen
t
sen
so
r
a
nd
com
m
un
ic
at
io
n
syst
e
m
can
be
sm
e
ared
i
n
ag
ricul
tural
m
on
it
or
i
ng
a
nd
co
ntr
ol
.
Dee
p
neural
netw
ork
(
D
NN
)
tr
ai
ni
ng
c
onta
ine
d
appr
ox
im
at
ely
1300
pict
ures
an
d
was
i
ncre
m
ented
w
he
n
m
or
e
insect
s w
ere
tr
app
e
d
f
or
the
pe
rio
d
of
the
ini
ti
al
te
sti
ng
.
C
odli
ng
m
oth
an
d
ge
ner
al
insect
s
we
re
t
wo
cl
a
sses o
f
the
dataset
i
n
a
dd
it
io
n
detect
ed
obj
ect
is
t
o
a
ge
ner
al
i
ns
ect
or
the
ta
r
get
C
od
li
ng
M
oth
pr
ov
i
ded
by
D
N
N
[
9
].
Sara
nya
et
al
.
i
n
their
resea
rc
h,
a
co
ntro
ll
in
g
syst
e
m
fo
r
pe
st
wh
ic
h
colla
borated
the
exi
ste
nce
of
pests
in
the
far
m
ing
la
nd
thr
ough
Pas
siv
e
infrare
d
se
nsor
a
nd
im
age
processi
ng
m
eth
od
wh
ic
h
pro
du
ce
s
ultras
ou
nd
t
hat
was
ins
uffer
a
bl
e
to
ro
de
nts
and
inse
ct
s
[10].
In
the
ag
ricul
tural
m
on
arch
y
,
detect
ing
cr
op
disease
an
d
pests
is
a serio
us
d
if
fic
ulty
. Tr
aditi
on
al
p
est
d
et
ect
io
n
proce
dures
a
re d
if
ficult
, tim
e
-
co
nsum
ing
, an
d pro
ne
to m
i
sta
ke.
In
recent
ye
ars
,
there
has
be
e
n
a
great
er
f
oc
us
on
re
searc
h
stud
ie
s
on
t
he
us
e
of
va
rio
us
strat
egies
in
th
e
fiel
d
of ag
ricult
ural
pest m
anag
em
ent.
3.
SY
STE
M AR
CHI
TE
CT
U
R
E AND
M
ODE
L
In
our
st
udy,
we
se
par
at
e
our
m
ai
n
m
od
el
into
tw
o
i
nd
e
pende
nt
sub
s
ect
ion
on
e
is
t
he
propose
d
m
od
el
an
d
anot
her
one Syste
m
A
rch
it
ect
ur
e
. F
or
the
Io
T
de
vice stru
ct
ure
m
easur
em
ent o
r
desig
n
m
od
el
ing
of
the
de
vice
is
ver
y
undoubte
dl
y
essenti
al
to
go
to
the
ne
xt
pr
ocedu
re.
N
owa
days,
de
velo
pm
ent
in
Io
T
[
11]
an
d
trackin
g
syst
e
m
has
so
lve
d
our
co
nventi
onal
an
d
eg
reg
i
ou
s
pro
blem
s
m
or
e
eff
ic
ie
nt
way.
For
Ac
cur
at
e
m
od
el
ing
an
d
execu
ti
on,
we
hav
e
desig
ne
d
two
A
rc
hitec
tural
m
od
el
s
so
that
any
inco
nsi
der
a
ble
ad
dress
of
m
od
el
ing
poi
nt can
e
vid
e
ntly
be
ide
ntifie
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
IFSG
:
I
ntell
ige
nce
agric
ulture
crop
-
pest
dete
ct
ion
syste
m us
ing
I
oT
auto
m
ation syst
em
(
I
mrus
Sa
le
hin
)
1093
3.1.
Pr
opose
d model
In
this
sect
io
n,
we
ha
ve
des
ign
e
d
an
a
utom
at
ed
syst
e
m
that
is
set
up
in
the
A
gr
i
fi
el
d.
I
n
T
his
syst
e
m
,
we
are
us
ing
four
ba
sic
adv
a
nced
sensors
[
12
]
,
[1
3]
f
or
pe
st
detect
ion
an
d
m
on
it
or
in
g.
Fo
r
the
autom
at
ion
pr
ocess,
we
e
sta
blishe
d
a
GSM
m
od
ule
a
nd
al
so
W
i
-
Fi
c
onnecti
vity
.
Our
m
ai
n
ta
rg
et
is
SMS
al
erts
to
t
he
e
nd
-
us
er
or
Far
m
er
ab
ou
t
the
fiel
d
sit
u
at
io
n.
I
n
Fig
ur
e
1
w
e
set
up
al
l
com
po
ne
nts
in
to
th
e
Ardu
i
no
dev
ic
e
. T
his m
od
el
is v
e
ry
ad
va
nce
d
as
w
el
l as l
ow c
os
t.
Figure
1. De
vi
ce o
rg
a
nized
pro
ces
s m
od
el
3.2.
S
ystem
ar
chitecture
In
this
syst
e
m
as
sh
own
in
Fi
gure
2
,
we
us
e
d
fou
r
cheap
a
nd
highly
sensiti
ve
senso
r
s
to
m
on
it
or
,
detect
an
d
pr
e
ven
t
pest
at
ta
c
ks
i
n
the
ag
ric
ultur
al
fie
ld.
By
us
in
g
t
hese
f
our,
we
m
ade
an
a
uto
m
at
ed
dev
ic
e
that
can
analy
ze
data
fr
om
pr
evi
ou
sly
store
d
an
d
pro
gr
am
m
ed
by
Ar
dui
no
Uno
R3
an
d
ha
ve
the
abi
li
ty
to
sen
d
data
to
use
rs
ab
ou
t
rea
l
-
tim
e
pest
inf
or
m
at
ion
of
hi
s
fiel
d
with
ha
ving
a
long
-
distance
us
i
ng
GS
M
SI
M0
0A Mo
du
le
.
Figure
2. Ci
rcui
t desig
n
f
or IF
SG
a
uto
m
at
ion
syst
e
m
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
2
4
, N
o.
2
,
N
ove
m
ber
20
21
:
10
9
1
-
109
9
1094
We
ha
ve
si
x
ha
rdwar
e
i
n
ou
r
syst
e
m
tho
se
are
deci
bel
det
ect
ion
se
ns
or
HH0
6.03,
ther
m
al
senso
r
us
in
g
AM
G88
33,
cam
era
mo
dul
e
0v76
70,
gas
detect
ion
sensor
MQ
-
138,
Ard
uino
Uno
R3
an
d
GS
M
SI
M9
00A
m
odule
.
I
n
this
Syst
e
m
,
we
us
e
d
four
chea
p
a
nd
hi
gh
ly
se
ns
it
ive
sens
ors
to
m
on
it
or
,
detec
t
an
d
pr
e
ve
nt
pest
at
ta
cks
in
the
a
gr
ic
ultur
al
fiel
d.
By
us
i
ng
th
ese
fou
r,
we
m
ade
an
a
utom
at
ed
dev
ic
e
t
hat
can
analy
ze
data
from
pr
evio
us
ly
store
d
an
d
pro
gr
a
m
m
ed
by
Ar
duin
o
U
no
R3
and
ha
ve
the
a
bili
ty
to
send
da
ta
to
us
ers
a
bout
rea
l
-
tim
e
pest
info
rm
at
ion
of
hi
s
fiel
d
with
ha
ving
a
lon
g
-
distance
us
in
g
GSM
SI
M00A
Module
.
We
hav
e
six
hard
war
e
i
n
our
syst
e
m
those
are
decibel
detect
ion
sen
so
r
H
H
06.03,
therm
al
sens
or
us
i
ng
AMG
8833,
ca
m
era
m
od
ule
0v7670,
gas
de
te
ct
ion
sen
sor
MQ
-
138,
A
rduin
o
U
no
R3
and
GS
M
SIM
900A
Module
.
Sound
s
e
nsor:
We
kn
ow
t
hat
pests
are
us
in
g
30dB
to
52dB
for
their
i
nter
nal
com
m
un
ic
at
ion
.
Pe
sts
use
to
at
tract
their
opposit
e
sex
ua
l
m
aking
noise
betwee
n
20
0Hz
to
60
0H
z
[
14
]
.
He
re
we
us
e
d
a
se
nsor
t
o
cat
ch
a
nd
anal
yz
e
their
s
ound.
A
decibel
de
te
ct
ion
m
od
ul
e
sou
nd
sen
sor
with
s
erial
tra
ns
ist
or
-
tra
ns
ist
or
log
ic
(
T
TL
)
outp
ut
ha
ving
30dB
to
130d
B
and
40Hz
-
8
kH
z
Sens
it
ivi
ty
is
us
ed,
it
s
m
od
el
nam
e
is
HH_
06.03.
It’s
so
c
hea
p.
As
i
t
can
be
op
e
rated
in
5V
it
’s
s
o
easy
to
co
nn
ect
it
with
the
Ardu
i
no
U
no
R3
bo
a
r
d.
Its
VCC
a
nd
GND
ar
e
connecte
d
t
o
Ardu
i
no’s
5V
and
GND.
H
H_0
6.03’s
‘
T
X’
a
nd
‘RX
’
Pin
will
b
e c
onnect
ed
to
TX
and
RX of
Ard
uino
Uno
R
3
as t
hi
s sen
s
or h
a
ve
t
ran
sist
or
-
tran
sist
or
l
og
ic
(TT
L)
ou
t
pu
t.
It
will
analy
ze
the
am
bient
noise
an
d
will
identify
the
sound
as
pro
gr
am
m
ed
in
a
sp
eci
fic
ra
nge
as
It
is
pests
so
un
d
or
no
t.
It
will
identify
the
noise
of
pest
wi
ngbeats
or
s
ound
by
their
m
ou
th
of
anyt
hing else
f
or their
basic i
ns
ti
nct [1
5].
Ther
m
al
s
ensor:
This
se
ns
or
is
us
ed
to
se
ns
in
g
fl
uoresc
ence
[
16
]
.
C
hlo
r
ophyll
flu
ore
scence
will
be
identifie
d
by
th
is
sensor
as
a he
al
thy
le
af’
s
fl
uoresce
nce
dat
a
will
be
store
d
pre
viously
.
By
analy
zi
ng
th
e
store
d
an
d
real
-
tim
e
pictorial
view
the
Ardu
i
no
will
say
as
pro
gr
am
ed
whet
her
it
is
at
ta
c
ked
by
pests
or
no
t.
He
re
we
us
e
d
the
AMG
883
3
T
her
m
al
sens
or
m
odule.
Its
SCL
a
nd
SDA
Pin
a
re
c
onnected
to
A
5
and
A
4
i
n
Ard
uino
Uno
c
ons
ecuti
vely
.
VC
C
an
d
G
N
D
a
r
e
co
nn
ect
e
d
t
o
5V
a
nd
G
N
D
in
the
A
r
du
i
no
Uno b
oard.
Cam
era
m
od
ule:
Her
e
we
use
d
a
CM
OS
O
V76
70
Ca
m
era
Mod
ule
1/
6
-
In
c
h
0.3
-
Me
ga
pix
el
Module
to
identify
pests.
This
cam
era
mo
dule
will
sens
e
the
m
ajo
r
ch
ang
e
in
it
s
i
m
a
ge
[17].
This
c
a
m
era
m
od
ule
has
an
im
age
arr
ay
capa
ble
of
op
e
rati
ng
at
up
t
o
30
f
ram
es
per
sec
ond
(
FPS
)
in
VGA
with
com
plete
us
er
c
ontrol
over
im
age
qu
a
li
ty
,
fo
rm
atting
an
d
ou
t
pu
t
da
ta
trans
fer.
All
require
d
im
age
processi
ng
functi
ons,
i
nclud
i
ng
ex
posur
e
co
ntro
l,
gam
m
a,
wh
it
e
balance,
c
olor
sat
urat
ion,
hue
c
ontr
ol
an
d
m
or
e
,
are
al
so
pro
gra
m
m
able
throu
gh
the
SCC
B
i
nterf
ace
.
It'
s
so
easy
to
co
nn
ect
with
Ardu
i
no
U
no
R3
.
T
he
ca
m
era m
od
ule
is p
rope
rly
conn
ect
e
d wit
h A
rduin
o U
no R3
as a
giv
e
n
ci
r
c
uit diag
ram
.
Gas
d
et
ect
io
n
s
ens
or
:
A
n
M
Q
-
138
Gas
de
te
ct
ion
sens
or
is
us
ed
i
n
thi
s
autom
at
ed
syst
e
m
.
It
has
a
wides
pr
ea
d
tra
ckin
g
sc
op
e
,
r
apid
res
pons
e
and
stron
g
se
ns
it
ivit
y,
fixe
d
and
lo
ng
li
fe
and
ha
ving
a
si
m
ple
dr
ive
ci
rcu
it
.
It
is
use
d
in
Breat
h
al
co
ho
l
detect
ors,
s
olv
e
nt
detect
ors,
Ai
r
qu
al
it
y
con
t
ro
l
ty
pes
of
equ
i
pm
ent
fo
r
bu
il
di
ngs/offi
c
es.
It’s
VCC,
DOUT,
A
OU
T
,
GND
Pi
n
co
nnect
ed
to
5V,
A1,
A
0,
GND
in
Ardu
i
no
U
no
R3
Boar
d.
Et
hyle
ne,
nitric
ox
i
de,
j
asm
on
ic
,
m
et
hyl
j
asm
on
at
e,
oci
m
ene,
li
m
on
ene
,
plasto
qu
i
none,
ge
ran
i
ol,
li
nal
oo
l,
ci
tro
nellol
,
an
d
ly
co
pe
ne
are
dif
fer
e
nt
VO
Cs
that
c
om
e
fr
om
plant
s.
This se
nsor
w
il
l i
den
ti
fy s
om
e
of th
os
e c
om
po
un
ds
a
nd ide
nt
ify
the h
eal
th
of the c
rop.
Ardu
i
no
U
no
R3
an
d
G
SM
SI
M9
00A
M
odule:
I
n
this
pro
j
ect
,
Ardu
i
no
U
no
R3
is
us
e
d
to
analy
z
e
colle
ct
ed
data
from
tho
se
f
our
se
nsor
s.
S
om
et
i
m
es
it'
s
a
naly
sis
data
th
at
is
pr
e
viousl
y
stored
by
th
e
pro
gr
am
.
On
t
he
oth
er
ha
nd,
GS
M
S
IM9
00
A
Mo
dule
is
use
d
to
sen
d
dat
a
at
the
us
e
r
e
nd.
GS
M
M
odul
e
can se
nd
data
with a l
ong
-
dis
ta
nce and a
t an
y si
tuati
on
li
ke
d
e
ns
e
fog o
r h
eavy rai
n.
4.
METHO
DOL
OGY
Fo
r
the
researc
h
stu
dy,
we
as
so
ci
at
e
al
l
m
ater
ia
ls
com
po
se
d
of
a
m
ic
ro
co
ntr
oller
an
d
dif
fer
e
nt
ki
nds
of
a
dv
a
nce
d
se
ns
ors
li
ke
I
FS
G
sens
or,
G
S
M
m
od
ule,
cr
op
an
d
sen
sor
colla
borati
on.
A
ll
data
fr
om
th
e
GS
M
Module
w
ou
l
d
be
receive
d
on
a
pest
detect
ion
de
vice
an
d
the
Ard
uino
w
ou
l
d
al
so
se
nd
the
regulat
in
g
act
ion
to the c
hip
by
eff
ic
ie
nt C++
pro
gr
am
m
ing
.
4.1.
G
SMSI
M90
0A m
od
ul
e st
ruc
tu
r
al se
tu
p
The
SI
M
900A
is
an
esse
ntial
an
d
un
i
qu
e
G
SM/
GP
RS
m
od
ule
us
e
d
i
n
diff
e
ren
t
kinds
of
I
oT
fiel
ds.
The
m
od
ule
can
al
so
be
util
ized
to
de
velo
p
I
oT
an
d
Em
bed
ded
Applic
at
io
ns
.
It
w
orks
in
the
900
-
1800
MHz
fr
e
qu
e
ncy
ra
nge.
An
RS
232
interface
is
incl
ud
e
d
with
the
m
od
e
m
,
al
lowin
g
yo
u
to
co
nn
ect
a
PC
as
we
ll
as
a
m
ic
ro
co
ntro
ll
e
r
with
a
n
RS
23
2
chi
p
(M
AX2
32).
T
he
i
nbuilt
TCP/
IP
sta
ck
in
the
GS
M/
G
PRS
m
od
em
a
l
lows
you
t
o
c
onnec
t
to
the
inter
ne
t
thr
ough
GPR
S.S
IM
900A
GS
M/
GP
R
S
Mod
em
Feat
ures:
(
1)
I
nput
Vo
lt
age
:
12V
DC
(
2)
S
upports
MIC,
Audio
I
nput
&
Sp
eake
rs
(
3)
Du
al
-
Ba
nd
GSM
/GPRS
900/
1800
MHz
(4)
RS232
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
IFSG
:
I
ntell
ige
nce
agric
ulture
crop
-
pest
dete
ct
ion
syste
m us
ing
I
oT
auto
m
ation syst
em
(
I
mrus
Sa
le
hin
)
1095
interface
sel
ec
ti
on
(5)
Lit
hium
batte
ry
In
te
rf
ace
(
6)
Co
nfi
gurab
le
ba
ud
rate
(7)
A
ntenn
a
(
SMA
ant
enna
interfaces
)
(
8)
SI
M C
ar
d
slot
(9)
Net
wor
k
St
at
us
LE
D
(B
uilt
-
in) (10)
B
uilt
-
in
powerf
ul TCP/IP pro
t
oc
ol stack
for
GP
R
S
inter
net
data
trans
fe
r
(11
)
D
ATA
G
PRS:
do
wn
l
oa
d
trans
fe
r
m
a
x
is
85.6KB
ps
,
Up
loa
d
tra
ns
f
er
m
ax
42.8KBp
s.
In
Figure
3
s
hows
the
fu
ll
stract
ure
of
t
his m
od
ul
e stru
ct
ural
set
up.
Figure
3. G
SM SIM
900A
p
r
oto
col
Connect
io
n
set
up
:
no
wa
days
,
we
interface
a
GS
M
m
od
ul
e
with
an
A
rduin
o
m
od
ule
to
sen
d
data.
The
a
dvanta
ge
of
the
G
SM
m
odule
si
gn
al
is
avail
able
in
a
wide
ra
ng
e
of
areas.
GS
M
m
odule
wor
ks
w
it
h
AT
com
m
and
s.
Us
ing
Io
T
co
nnec
ti
on
s,
we
hav
e
recorde
d
data
f
ro
m
a
far
m
er'
s
fiel
d.
In
t
his
a
rtic
le
,
we
are
goin
g
to
set
up
a
ci
rcu
it
diagr
am
of
Ardu
i
no
to
interfacin
g
the
G
SM
m
od
ule.
T
o
sen
d
sens
or
data
to
Ardu
i
no
U
no
R3
an
d
recei
ve
SMS
al
erts,
we
ha
ve
us
e
d
the
S
IMCO
M,
SI
M9
00A
-
GS
M
m
od
ule.
It’s
pr
et
ty
sim
ple
to
interfaci
ng w
it
h Ardu
i
no a
nd
GS
M m
od
ules
in Figu
re
4.
Figure
4. G
SM SIM
900A
s
tr
uc
tural fu
nctio
n
4
.
2.
Crop pes
t
f
unct
i
onal
analysis
In
sect
s
,
viruse
s,
an
d
f
ungi
at
ta
ck
s
om
eti
m
e
s
cause
a
huge
of
dam
age
in
the
ag
ricult
ura
l
sect
or
.
T
o
avo
i
d
this
pro
bl
e
m
,
we
wo
r
ke
d
on
4
sen
sors:
so
und
detect
io
n
sens
or,
flu
oresce
nce
sen
sor
,
high
-
im
age
sens
or
,
and
ga
s
se
nsor
.
As
a
res
ult
o
f
these
4
se
ns
ors,
a
ny
disease
can
be
detect
ed
befor
e
the
f
ie
ld
is
dam
aged
a
nd
there
is
an
op
port
un
it
y
to
ta
ke
necessary
act
ion
acc
ordin
gly.
W
it
h
the
i
m
pro
vem
ent
of
sci
ence,
these
s
ens
or
s
will
n
ow
open
a com
pr
ehe
ns
i
ve doo
r
f
or im
pro
vem
ent in ag
ric
ultur
e.
T
he
us
e
of
s
ou
nd
to
ide
ntify
insect
s
is
a
gro
undbrea
king
i
nventio
n.
T
hese
sens
ors
ha
ve
a
nten
nas
t
hat
ar
e
able
to
receive
var
i
ou
s
sou
nds
from
insect
s.
Most
insect
s
usual
ly
m
ake
a
so
un
d
of
4000
Hz
to
2000
0
H
z
range.
O
ur
sen
so
r
recei
ved
al
l
sounds
in
t
he
r
ang
e
bet
ween 400
0
-
2000
0
H
z.
E
ver
y
ins
ect
create
s
it
s ow
n
ty
pe
of
s
ound.
And
the
f
re
qu
e
ncy
diff
e
rs
f
r
om
each
oth
er.
Usu
al
ly
,
the
E
ns
ife
ra
sub
-
ord
er
of
Or
t
hopter
a
order
a
nd
the
Ci
cado
idea
s
ub
-
fam
i
ly
of
H
om
op
te
ra
orde
r
us
e
d
s
ound
f
or
m
at
ing
,
Co
pu
la
ti
on
in
du
c
es
an
d
te
rr
it
ori
es
are
decla
red
a
nd
de
fe
nd
e
d,
an
d
ot
her
pur
pos
es
[18].
Genus
gr
yl
lus
c
ricket
s
create
s
ounds
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
2
4
, N
o.
2
,
N
ove
m
ber
20
21
:
10
9
1
-
109
9
1096
by scr
a
ping a s
crap
e
r
on a fo
r
ewin
g
al
ong
it
s
o
pp
os
in
g
f
or
e
wing for
rou
gh teet
h
and
t
he
r
at
e is increased
by
the
risin
g
of
te
m
per
at
ure
[19].
Be
ca
us
e
of
t
he
cl
ean
and
lo
w
fr
e
quency,
they
c
re
at
e
a
eu
phonic
so
un
d.
Th
rou
gh
the
lo
w
-
fr
e
quency
s
ound
a
nd
the
s
ound
kind,
se
nsor
s
detect
them
.
Ci
cada
(Cic
ad
et
ta
m
on
ta
na)
is
an
al
ar
m
ing
pest
for
the
w
oody
plant
w
hich
m
akes
w
ho
le
in
the
ste
m
s.
An
d
eggs
la
y
inside
the st
em
s [
20
]
.
Th
ey
pro
du
ce
a h
ig
he
r
f
re
qu
e
ncy s
ound
wh
i
ch
is l
ess
pur
e
than Ge
nus
gr
y
ll
us
.
Chloro
ph
yl
l
fluoresce
nce
se
ns
or
us
e
d
a
la
ser
or
LE
D
li
gh
t
that
evaluates
it
by
tr
ansf
e
rr
i
ng
the
photosy
nth
et
ic
el
ect
ron
[
21]
.
Wh
ic
h
is
vis
ua
li
zed
the
e
ff
ic
ie
ncy
of
photosy
nth
esi
s.
Pict
ur
es
of
healt
hy
gr
ai
n
le
aves
ar
e
al
read
y
inp
ut
te
d
in
the
database
of
the
sens
or
.
I
f
the
le
af
is
infested
with
fungus
or
insect
s,
this
ti
m
e
the
patte
r
n
of
c
hloro
ph
yl
l
is
co
m
pared
with
the
previ
ou
s
one.
The
c
hange
i
n
chlo
rop
hyll
can
be
see
n
on
t
hat
le
af.
A
nd
a
ccordin
g
to
the
change
in
t
he
chlo
rop
hyll
of
the
le
aves,
it
is
easy
to
find
out
wh
at
ki
nd
of
f
ungus
or
insect
has
at
ta
cked.
Ma
ny
ins
ect
s
(A
no
plo
c
nem
is
cur
vip
e
s,
Haly
om
or
pha
haly
s,
an
d
Schi
zaph
is
gr
am
inu
m
)
eat
the
le
aves
of
t
he
cr
op
an
d
m
any
insect
s
change
t
he
le
aves
(b
y
la
yi
ng
e
gg
s
)
in
a
diff
e
re
nt
way.
Si
m
i
la
rly
,
the
sensor
detect
s
the
var
io
us
ty
pes
of
fun
gus
(P
uc
ci
nia
recond
it
e,
P
hytoph
thora
infesta
ns,
Cochlio
bolus
hetero
st
rop
hus,
Bi
po
la
ris
m
ay
dis,
Erw
i
nia
am
yl
ov
or
a
,
an
d
Xa
nthom
on
a
s
or
yz
ae
)
by
var
i
ou
s
s
pots
unde
r
or
a
bove
the
le
aves.
Leaf
spot,
Lea
f
Bl
igh
ts, R
us
ts,
Powd
e
ry M
il
de
w,
a
nd
Dow
ny
Mi
ldew
are
the m
ajo
r
funga
l diseases.
Each
plant
re
flect
s
a
certai
n
a
m
ou
nt
of
li
ght
ene
rg
y
i
n
nat
ur
e
.
Using
t
his
ref
le
ct
io
n
c
on
cept,
a
n
overal
l
idea
ab
out
the
gr
ai
n
can
be
obta
ined
thr
ough
this
high
-
im
age
se
nsor.
Pla
nts
di
ffuse
d
di
ff
e
ren
t
ty
pes
of
rays
su
c
h
as
x
-
rays,
U
V,
in
f
ra
red,
an
d
ra
dio
wav
e
s
.
T
he
re
f
le
ct
ed
sp
ect
ral
sign
at
ur
e
data
from
each
crop
is
pre
-
recorde
d
i
n
the
da
ta
ba
se
by
a
high
-
i
m
age
sens
or.
Wh
e
n
the
c
r
op
is
at
ta
c
ked
by
a
n
ins
ect
or
disease,
the
am
ount
of
re
flect
ed
li
gh
t
is
ch
a
ng
e
s.
T
hen
the
data
is
com
pa
red
with
previ
ou
s
data
with
t
he
help
of
data
m
ining
,
w
hich
becam
e
a
vit
al
te
chn
ol
og
y
in
plant
sci
en
ce
[22].
This
giv
es
a
n
idea
of
wh
et
her
t
he
cr
op
is
disease
d
or
i
nf
ect
e
d
by
insect
s.
F
or
exam
ple,
a
he
al
thy
bar
le
y
le
af
re
flect
s
ab
out
0.6%
of
the
li
gh
t
of
1000
nm
wav
el
en
gth
,
bu
t
w
he
n
the
le
af
is
aff
ect
ed
by
r
us
t
dis
ease,
the
sam
e
wav
el
e
ng
t
h
de
creases b
y
0.5
%. I
n
t
he
case of
pow
der
y m
i
ldew disea
ses,
the r
eflect
io
n
is furthe
r
re
du
c
ed
to
ab
out
0.4
5%
.
G
rain
s
f
or
diff
e
re
nt
disea
ses
gi
v
e
diff
e
r
ent
re
flect
ion
s
at
the
sam
e
web
le
ngth.
F
r
om
this, it
is facil
e
to g
et
a
n
i
dea a
bout the
att
ack
of v
a
rio
us
dise
ases th
rou
gh th
is sens
or.
A
plant
ca
n
pro
duce
m
or
e
t
han
one
la
kh
chem
ic
al
co
m
pounds.
A
bout
1700
chem
icals
are
vola
ti
l
e
a
m
on
g
them
[2
3].
P
la
nts
use
d
vola
ti
le
org
anic
com
pounds
to
de
fend
t
hem
sel
ves
aga
inst
insect
s,
for
po
ll
inati
on,
f
or
com
m
un
ic
at
ion
betwee
n
plants
them
sel
ves
[24].
In
ad
diti
on,
if
a
pla
nt
is
infecte
d
by
a
path
og
e
n,
it
produce
d
a
vola
t
il
e
co
m
pound.
In
t
he
sen
sor,
sam
ples
of
any
vola
ti
le
su
bs
ta
nce
em
it
ted
from
a
plant
unde
r
any
ci
r
c
um
sta
nces
are
inputt
ed
i
nto
the
data
from
befor
e
.
Wh
e
ne
ver
t
he
se
nsor
detect
s
su
c
h
ty
pe
of
s
ubsta
nc
e,
it
pr
ese
nte
d
the
co
ndit
ion
of
t
he
plants/c
rops
by
recei
vi
ng
t
he
vola
ti
le
su
bst
ance.
Cuc
um
ber
,
ca
bba
ge
,
co
rn,
tom
at
o,
an
d
li
m
a
bean
s
plant
releas
es
a
chem
ic
al
t
hat
at
tract
s
the
pr
e
dato
rs
of
he
rb
i
vores.
T
he
grow
t
h
an
d
ge
rm
inati
on
of
Mon
il
ina
la
xa
are
co
ns
ide
ra
bly
reduce
d
by
carv
ac
ro
l,
tra
ns
-
2
-
he
xe
nal,
a
nd
ci
tral
w
hic
h
com
e
fr
om
plants
[
25]
.
Ethyl
ene,
nitr
ic
ox
i
de,
j
asm
on
ic
,
m
et
hyl
j
as
m
on
at
e,
ocim
ene,
l
i
m
on
ene,
plastoquin
one,
ger
a
nio
l,
li
nalo
ol,
ci
tron
el
lol,
a
nd
ly
cop
e
ne
ar
e
diff
e
re
nt
V
OC
s
that
c
om
e
from
plants.
T
he
gas
se
ns
or
det
ect
ed
this
ty
pe
of
VO
Cs
an
d
easi
ly
detect
e
d
al
l st
ress
of
pla
nts.
4.
3.
Se
ns
or
i
nt
e
gration
wi
t
h GSM
In
t
he
pr
ece
ding
sect
io
ns
,
t
he
syst
e
m
pr
oto
ty
pe
is descri
bed
pro
per
ly
.
Se
nsors
are u
se
d
to
r
ecei
ve
t
he
data
f
ro
m
the
env
i
ronm
ent
and
t
he
processi
ng
m
et
hodo
l
ogy
is
done
by
A
rduin
o
Uno
R
3.
A
fter
process
data
from
sensors,
Ardu
i
no
Uno
pr
e
par
e
d
t
he
final
res
ult
as
A
rduin
o
is
pro
gra
m
m
ed
in
be
f
or
e
.
W
it
h
the
help
of
the
G
SM
SI
M
900A
m
od
ule,
this
final
res
ul
t
is
delivere
d
thr
ough
this
sy
stem
us
er’
s
m
ob
il
e
phone
as
a
t
ext
SMS.
Usi
ng
t
his
syst
e
m
us
er
/farm
ers
can
rem
otely
sense
their
file
d
real
-
ti
m
e
con
di
ti
on
.
This
sy
stem
introd
uced
GSM
to
trans
fer
da
ta
so
it
co
uld
be
use
d
from
m
il
e
to
m
il
e.
In
this
researc
h
four
sens
ors
ar
e
us
e
d.
It
is
po
s
sible
t
o
tra
ns
m
it
fo
ur
sens
ors’
analy
zed
data
to
the
us
e
r'
s
cel
l
ph
on
e
bot
h
in
divi
du
al
ly
an
d
t
oget
he
r.
Sens
or
'
s
sta
tus
cou
ld
be
che
cked
by
sen
din
g
a
te
xt
SM
S
to
GS
M
N
o.
of
S
IM9
00A
Module
.
I
n
an
au
di
o
detect
ion
sens
or,
the
sou
nd
m
ade
by
pests
(e.
g.,
wi
ng
be
at
s,
an
d
at
tra
ct
ano
t
her
sex
ually
)
is
detec
te
d.
A
sp
eci
fic
ra
ng
e
of
decibels
is
us
e
d
for
sp
eci
f
ic
pests.
By
us
ing
this
sen
sor
with
pro
per
A
rduin
o
pro
gra
m
pests
are
identifie
d.
And
to
sen
d
this
data
rem
ot
el
y
to
the
field
owne
r
GS
M
m
od
ule
is
need
ed
.
The
re
st
of
t
he
sens
or
s
ha
ve
t
he
sam
e
char
a
ct
erist
ic
s
of
re
cei
vin
g
or
c
ollec
ti
ng
data
f
rom
the
am
bient
en
vir
on
m
ent
and
ar
e
analy
zed
via
Uno
R3
.
I
n
thi
s
syst
e
m
,
we
m
ade
a
strong
and
a
uto
m
at
ed
Ardu
i
no
pro
gram
with
wh
ic
h
if
the
four
sen
sors
prov
i
de
a
n
unusu
al
read
i
ng,
on
e
te
xt
SMS
will
sen
d
th
e
use
r'
s
cel
l
phone
a
uto
m
atical
ly
.
But
in
a
regular
sit
uatio
n
use
rs
ha
ve
to
sen
d
a
te
xt
S
MS
in
the
S
IM
900A
m
od
ule
t
o
m
on
it
or
th
e
s
ens
or
r
eadi
ng
a
s
well
as the a
gr
ic
ultu
ral fiel
d from
pest
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
IFSG
:
I
ntell
ige
nce
agric
ulture
crop
-
pest
dete
ct
ion
syste
m us
ing
I
oT
auto
m
ation syst
em
(
I
mrus
Sa
le
hin
)
1097
5.
E
X
PERI
MEN
TAL A
N
ALY
SIS
In
this
sect
io
n
i
m
plies
that
the
m
ai
l
auto
m
ation
syst
em
vis
ualiz
at
ion
an
d
dem
on
strat
io
n
in
real
li
fe.
We
use
this
de
vice
f
or
real
-
ti
m
e
data
colle
ct
ion
an
d
a
naly
sis
about
it
s
w
orkin
g
m
echani
s
m
.
At
the
fir
st,
it
is
set
u
p wit
h
al
l sens
or
s i
nto
a
hard
woo
d
then
the f
ull sy
ste
m
instal
ls i
nto
the f
ie
ld. In
Fi
gure
5, sho
ws visuali
ze
the pr
op
e
r func
ti
on
al
ar
ra
ng
e
m
ent o
f
this a
ut
om
a
ti
on
syst
e
m
.
Figure
5. F
ull
s
et
up
syst
e
m
f
or r
eal
-
li
fe
dem
on
st
rati
on
In
this
ci
rcu
it
bo
a
r
d,
we
us
e
al
l
pr
em
iu
m
and
lo
ng
-
la
sti
ng
se
nsors
wit
h
so
m
e
reg
ist
er
and
L
E
D
disp
la
y.
All
th
e
f
un
ct
io
ns
co
nn
ect
with
the
GS
M
m
odule
f
or
the
data
ca
r
d
a
nd
pe
st
dete
ct
ion
protoc
ol.
From
the ex
am
inati
o
n,
w
e create
a m
od
el
ch
at
an
d
data ta
ble w
hich
is used
for
re
su
lt
an
al
ysi
s.
Ardu
i
no U
no
R3 and
GS
M
S
IM9
00
A
Mo
dule
an
d
al
l
sensor
c
om
bin
at
ion
s
ha
ve
been
us
e
d
for
the
te
e
n
pe
st
and
m
od
er
at
e
pest
detect
ion Ot
he
rside
al
l
th
erm
al
sen
s
or
ss
us
e
d for all
criti
cal
p
est
detect
ion sy
stem
.
6.
RESU
LT
S
AND DI
SCUS
S
ION
S
To
the
outp
uts
ge
ner
at
e
a
nd
decisi
on
m
aking
we
ha
ve
se
t
up
a
syst
em
data
sh
e
et
.
I
n
t
his
Ta
ble
1,
i
m
plies
that
som
e
diff
e
ren
t
c
onditi
ons
a
nd
wh
ic
h
ty
pes
of
res
ult
outp
ut
from
the
auto
m
at
ion
syst
em
.
A
fter
analy
zi
ng
al
l
da
ta
,
we
com
pa
re
it
to
the
benchm
ark
sta
nd
e
r
data
evaluati
on.
W
e
us
e
I
S
O
an
d
IEEE
sta
nd
a
r
ds
value. A
fter t
he
set
-
up
de
vice
s,
it
works
92
% accu
rate
data stor
i
ng and
pro
vid
es
the
pe
r
fect SMS t
o
th
e u
se
r.
Tab
le
1
.
T
est
in
g
d
at
a
c
ha
rt
Para
m
eter
Test Co
n
d
itio
n
s
So
u
n
d
Sens
o
r
MQ13
8
CMOSOV7
6
7
0
AMG88
3
3
S
1
S
2
M
1
M
2
C
1
C
2
A
1
A
2
T
A
=2
5
°C
8
6
.9
8
9
.7
7
7
.2
89
8
8
.2
9
0
.1
8
7
.3
8
9
.2
Accurac
y
T
A
= 35
°C
9
0
.56
9
0
.78
8
0
.8
90
8
9
.2
9
0
.8
8
8
.9
9
0
.85
T
A
= 15
°C
8
2
.6
8
5
.8
7
5
.6
85
8
6
.8
7
7
.9
7
2
.6
9
8
.6
Sen
so
r
g
ain
T
MIN
≤
T
A
≤
T
M
AX
9
.9
m
V/°C
10
m
V/°C
8
.7
m
V/°C
9
.2
m
V/°C
Te
m
p
e
rature
co
ef
ficient
o
f
qu
iescen
t curre
n
t
4
0
°C ≤
T
A
≤ 14
°C
0
.39
0
.56
V
c
:5
.0V±0
.1V;
V
H
:5
.0V±0
.1V
0
.59
0
.62
Av
erage Resp
o
n
se rate
No
r
m
al
85%
88%
89%
82%
Test
Pe
rf
o
r
m
an
ce
No
r
m
al
Go
o
d
Go
o
d
Go
o
d
Go
o
d
Un
ite
°C
dB
△
Vs
f
p
s VGA
°C
In
this
data
c
har
t,
we
ha
ve
input
fou
r
di
ff
e
ren
t
ty
pes
of
se
nsor
valu
es
accor
ding
t
o
real
-
ti
m
e.
Condit
ion
ty
pe
is
the
m
ai
n
e
nv
i
ronm
ental
t
e
m
per
at
ure
(T
A
/°C
)
wh
e
n
w
e
te
st
the
senso
r
r
eadi
ng
a
nd
pes
t
detect
ion. F
or test
ing
se
nsor
dat
a, w
e
us
e
d
if
f
eren
t sam
ples li
ke
M
1
, M
2
/A
1
,
an
d
A
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
2
4
, N
o.
2
,
N
ove
m
ber
20
21
:
10
9
1
-
109
9
1098
7.
CONCL
US
I
O
N
The
inte
ntio
n
is
to
m
ai
ntain
up
with
t
he
la
te
st
sci
entifi
c
and
te
ch
nolo
gical
ad
va
ncem
ents
in
agr
ic
ultur
e
.
Cu
rr
e
ntly
,
there
i
s
no
al
te
rn
at
iv
e
to
sm
art
far
m
ing
to
produ
ce
m
or
e
cr
ops.
This
ne
cessi
ta
te
s
the
us
e
of
soph
ist
ic
at
ed
pest
cont
ro
l
te
chnolo
gy
.
Crop
outp
ut
is
ha
m
per
ed
the
m
os
t
by
these
insect
s.
With
this
pur
po
se
i
n
m
i
nd,
we
w
orke
d
us
in
g
four
se
ns
ors
t
o
achie
ve
this
go
al
.
Our
pest
-
rem
ov
al
au
t
om
at
ed
syst
e
m
s
will
be
a
n
i
ncredible
sci
e
ntific
su
cce
ss.
The
first
sens
or
is
the
s
ound
det
ect
ion
se
nsor
.
Thro
ugh
w
hic
h
th
e
no
ise
of
va
rio
us
so
rts
of
insec
ts
m
ade
in
the
fiel
d
is
identifie
d
an
d
data
is
sent
to
the
m
a
i
n
databa
se,
w
hi
le
the
ty
pe
of
insect
s
achin
g
is
al
s
o
detect
ed.
T
he
flu
or
esce
nce
s
ens
or
is
t
he
se
cond.
T
he
im
a
ge
of
t
he
le
af
will
be
colle
ct
ed
by
t
hi
s
sens
or,
w
hic
h
will
the
n
dis
play
the
am
ount
of
c
hlo
r
ophy
ll
in
that
le
af.
The
t
hird
is
th
e
high
-
i
m
age
sens
or,
wh
ic
h
will
be
capab
le
of
detect
ing
al
l
of
t
he
plant'
s
rays.
The
fou
rth
de
vice
is
a
gas
s
ens
or
,
wh
ic
h
can
dete
ct
al
l
gases
e
m
it
te
d
by
th
e
diseased
plant
.
Th
ere
was
a
centr
al
database
for
al
l
of
these
sen
so
rs
.
These
se
nsor
s
colle
ct
al
l
of
t
he
in
form
at
ion
an
d
c
om
par
e
it
to
the
ce
ntra
l
data.
Fa
rm
ers
can
e
nhance
gr
ai
n
yi
el
d
by
2
to
3
tim
es
by
ado
pt
ing
a
n
aut
om
ated
pest
rem
ov
al
syst
e
m
.
The
em
pl
oy
m
ent
of
four
dif
fer
e
nt
ty
pes
of
sen
sors
m
akes
detect
in
g
al
l
ty
pes
of
da
nge
rous
i
ns
ect
s
qu
it
e
si
m
ple.
Inse
ct
s
will
no
l
onger
be
a
s
co
urg
e
f
or
the
gr
ai
n.
As
a
res
ult,
the
a
gr
i
culture
in
dustr
y
m
a
y
underg
o
sig
nificant
c
ha
ng
e
s.
It
will
usher
in
a
new
era
in
a
gr
ic
ultur
e
.
ACKN
OWLE
DGE
MENTS
The
a
utho
r
ac
knowle
dges
the
su
pp
or
t
of
t
he
Pabna
U
niv
er
s
it
y
of
Scie
nce
and
Tec
hnolog
y
EEE
La
b
and
Da
ffo
dil
I
nter
national
U
niv
e
rsity
In
te
ll
igence
L
ab
a
nd
to
de
velo
p
t
his
aut
om
at
ion
de
velo
p
syst
em
.
Fo
r
bette
r
te
sti
ng, B
ang
la
des
h A
gri
culture
Uni
v
e
rsity
Facu
lt
y h
el
p
us as
well
.
REFERE
NCE
S
[1]
S.
Kim
,
M.
Le
e
,
and
C.
Shin,
“
IoT
-
Based
Stra
wberr
y
Dise
ase
Predic
ti
on
S
y
ste
m
for
S
m
art
Farm
ing
,”
Sensors
,
no.
18
,
no
.
11
,
2
018,
doi
:
10
.
339
0/s18114051
.
[2]
H.
Naga
r
and
R
.
S.
Sharm
a,
"A
Com
pre
hens
ive
Surve
y
on
Pest
Dete
c
ti
on
Te
ch
nique
s
using
Im
age
Proce
ss
ing
,
"
2020
4th
Inte
rnational
Confe
re
nce
on
Inte
llige
nt
Computing
and
Control
Syste
ms
(
ICICCS
)
,
2020,
pp.
43
-
48,
doi:
10
.
1109/ICI
CCS
48265.
2020.
9120889
.
[3]
C.
-
J.
Chen
,
Y.
-
Y.
Huang,
Y.
-
S.
Li
,
C.
-
Y.
Chan
g
,
and
Y
.
-
M.
H
uang,
"A
n
AIoT B
ase
d
Sm
art
Agric
ult
u
ral
S
y
ste
m
for
Pests Det
ec
t
i
on,
" i
n
IEEE Ac
ce
ss
,
vol
.
8
,
pp
.
180750
-
180761,
2020,
doi: 10.
11
09/ACCESS
.
2020.
3024891
.
[4]
A.
Archi
p,
N.
Bote
z
at
u,
E
.
Şerba
n,
P.
Herghe
le
giu
,
and
A.
Za
lă,
"A
n
IoT
b
ase
d
s
y
stem
for
remote
pat
ie
n
t
m
onit
oring,
"
2
016
17th
Int
ernati
onal
Ca
rpathian
Cont
rol
Confe
ren
c
e
(
ICCC)
,
2016,
pp.
1
-
6
,
doi:
10
.
1109/Ca
r
pat
hi
anCC.
2016.
7501056.
[5]
Y.
Li,
H.
W
ang,
L.
M.
Dang
,
A
.
Sadeghi
-
Ni
ara
k
i,
and
H.
Moon,
“
Crop
pest
r
ecogniti
on
in
n
at
u
ral
sc
ene
s
usin
g
convol
uti
on
al
neur
al
net
wor
ks
,”
Computers
and
El
ec
tronic
s
in
Agric
ult
ure
,
vol.
169,
2020,
doi:
10
.
1016/j.c
o
m
pag.
2019.
1051
74
.
[6]
M.
M.
Hasan,
I
.
Sale
hin
,
N.
N.
Moon,
T.
M.
K
a
m
ruz
za
m
an,
Ba
ki
-
Ul
-
Islam
,
and
M.
Hasan,
"A
Com
pute
r
Vision
S
y
stem
for
the
Cat
egor
izat
ion
of
Cit
rus
Frui
ts
Us
ing
Conv
olut
ional
Neura
l
Network,
"
2021
Inte
rnationa
l
Symposium on
E
le
c
tronic
s and
S
mar
t
Dev
ices (
ISESD)
,
2021,
pp
.
1
-
6,
doi
:
10
.
110
9/ISESD53023.2021.
9501790
.
[7]
R.
P.
L.
Durga
bai
,
P.
Bharg
av
i,
and
J
y
o
thi
S.
,
“
Pest
Mana
gement
Us
ing
Ma
chi
ne
L
ea
rn
ing
Algorit
hm
s:
A
Revi
ew
,”
Inte
r
nati
onal
Journ
al
of
Computer
Sci
enc
e
Eng
i
nee
ring
and
Information
Technol
ogy
Re
searc
h
(I
JCSE
ITR)
,
v
ol. 8,
no.
1,
pp.
13
-
22
,
2018
,
doi
:
10
.
24247/i
j
cse
i
trfeb20182
.
[8]
Z.
W
ang
,
K.
W
ang,
Z
.
L
iu,
X.
W
ang,
and
S.
Pan,
“
A
Cognit
ive
Vision
Method
for
I
nsec
t
Pest
Im
age
Segm
ent
at
ion
,”
I
FA
C
-
Pap
ersO
nLine
,
vo
l.
51
,
no.
17,
pp.
85
-
89
,
2
018,
doi
:
10
.
101
6/j
.
i
fac
o
l.
2018
.
0
8.
066
.
[9]
D.
Brunel
l
i,
A
.
Albane
se,
D.
d
'Acunto
and
M.
Narde
ll
o
,
"Ene
r
g
y
Neu
tra
l
Mac
hine
L
ea
rn
ing
B
ase
d
IoT
Dev
ice
fo
r
Pest
Dete
c
ti
on
i
n
Prec
ision
Agri
cul
tur
e,
"
in
I
EEE
Inte
rne
t
of
Thi
ngs
Magazine
,
v
ol.
2,
no
.
4,
pp
.
10
-
13,
Dec
ember
2019,
doi
:
10
.
11
09/IOTM.
0001.
1900
037
.
[10]
K.
Sara
n
y
a,
P.
Uva
Dhari
ni,
P.
Uva
Darshni
and
S.
Monisha,
"IoT
Based
Pest
Control
li
ng
Sy
stem
for
Sm
art
Agric
ult
ur
e,
"
2
019
Inte
rnation
al
Confe
renc
e
on
Comm
unic
ati
on
and
El
ec
tr
onic
s
Syste
ms
(
ICCES
)
,
2019,
pp.
1548
-
1552
,
doi:
10
.
1109/IC
CES45898.2019.
90
02046
.
[
1
1
]
N
.
M
a
t
e
r
n
e
a
n
d
M
.
I
n
o
u
e
,
"
I
o
T
M
o
n
i
t
o
r
i
n
g
S
ys
t
e
m
f
o
r
E
a
r
l
y
D
e
t
e
c
t
i
o
n
o
f
A
g
r
i
c
u
l
t
u
r
a
l
P
e
s
t
s
a
n
d
D
i
s
e
a
s
e
s
,
"
2
0
1
8
1
2
t
h
S
o
u
t
h
E
a
s
t
A
s
i
a
n
T
e
c
h
n
i
c
a
l
U
n
i
v
e
r
s
i
t
y
C
o
n
s
o
r
t
i
u
m
(
S
E
A
T
U
C
)
,
2
0
1
8
,
p
p
.
1
-
5
,
d
o
i
:
1
0
.
1
1
0
9
/
S
E
A
T
U
C
.
2
0
1
8
.
8
7
8
8
8
6
0
.
[12]
I.
Sale
h
in,
B
aki
-
Ul
-
Islam,
S.
M. N
om
an,
M.
M. H
asa
n,
S.
T
.
Dip
and
M.
Hasan
,
"
A Sm
art
Pollut
e
d
W
at
er
Overl
o
a
d
Drain
age
Det
ec
t
ion
and
Aler
t
Sy
stem:
Based
on
IoT,
"
2021
In
te
rnational
Mob
il
e
,
Inte
l
li
gen
t,
and
Ubiquit
ous
Computing
Conf
ere
nce (
MIUCC)
,
2021
,
pp
.
37
-
4
1,
doi
:
10
.
1109/
MIU
CC52538.2021.
9447664.
[13]
I.
Sal
ehi
n
,
I.
M
.
T
al
ha
,
M.
Sai
fuz
za
m
an
,
N.
N.
Moon
,
and
F.
N.
Nur,
"A
n
Advanc
ed
Met
hod
of
Tr
ea
t
ing
Agric
ult
ur
al
Cr
ops
Us
ing
Im
age
Proce
ss
ing
Al
gorit
hm
s
and
I
m
age
Data
Proc
essing
S
y
stems
,
"
2020
IEEE
5
t
h
Inte
rna
ti
onal
C
onfe
renc
e
on
Computing
Co
mm
unic
ati
on
a
nd
Aut
omat
ion
(
ICCCA)
,
2020,
pp.
720
-
7
24,
doi:
10
.
1109/IC
CCA49541.2020.
9250839
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
IFSG
:
I
ntell
ige
nce
agric
ulture
crop
-
pest
dete
ct
ion
syste
m us
ing
I
oT
auto
m
ation syst
em
(
I
mrus
Sa
le
hin
)
1099
[14]
Y.
Sale
h
,
M.
S.
Yah
y
a,
I
.
A.
D
aly
op,
and
R.
H
uss
ai
n,
“
W
ire
le
s
s
Sensor
Network
(W
SN
)
in
In
sec
t
Monitor
ing:
Acoustic
T
ec
hn
ique
in
I
nse
ct
Monitori
ng
(
A
Revi
ew/Surve
y
)
,
”
In
te
rnatio
nal
Journal
of
Engi
ne
ering
&
Technol
ogy
,
vo
l. 7, no. 3, pp. 121
-
126,
2018
,
doi
:
10.
14419/ijet.v7
i3.
36.
29091
.
[15]
I.
Potamiti
s
and
I.
Riga
kis
,
"N
ovel
Noise
-
Robust
Optoac
ousti
c
Sensors
to
Ide
nti
f
y
Insec
ts
Through
W
ingbe
at
s,"
in
IEE
E
Sensors
Jo
urnal
,
vol
.
15
,
n
o.
8
,
pp
.
4621
-
4
631,
Aug.
2015,
doi:
10
.
1109/JS
EN.
2015.
24249
24
.
[16]
A.
P.
Dem
che
nk
o,
“
Com
par
at
iv
e
Anal
y
sis
of
Flu
or
esc
ence
R
epor
te
r
Sign
al
s
Base
d
on
Int
ensity
,
A
nisotrop
y
,
Ti
m
e
-
Resolut
ion
,
and
W
ave
le
ngth
-
R
atiom
et
r
y
,”
A
d
v
a
n
c
e
d
F
l
u
o
r
e
s
c
e
n
c
e
R
e
p
o
r
t
e
r
s
i
n
C
h
e
m
i
s
t
r
y
a
n
d
B
i
o
l
o
g
y
,
vol.
8,
pp.
3
-
24
,
2010
,
doi:
10
.
1007/97
8
-
3
-
642
-
04702
-
2_1
.
[17]
O.
Lópe
z
,
M.
M.
Rac
h,
H.
Mi
gal
lon
,
M.
P.
Malumbres,
A.
Bonastre
,
and
J.
J.
Serra
no,
"M
onitoring
Pest
Insec
t
Tra
ps
b
y
Mea
ns
of
Low
-
Po
wer
Im
age
Sensor
T
ec
hnolog
ie
s
,
"
Se
nsor
s
,
vol.
12,
no.
11
,
pp.
15801
-
15819
,
2012,
doi:
10
.
3390/s12
1115801
.
[18]
G.
S.
Poll
ac
k
a
nd
K.
Im
ai
zumi
,
“
Neura
l
anal
ysis
of
sound
fre
quency
in
inse
c
ts
,”
Bi
oEssays
,
vol.
21
,
no
.
4,
pp.
295
-
303
,
19
99,
doi
:
10
.
1002
/(SICI)1521
-
1878(199904)21:4<295::AID
-
BIES5>3.
0.
CO;2
-
U
.
[19]
Brit
annica
,
Cr
ic
ke
t
,
Ency
c
lo
pedi
a
Br
it
ann
i
ca
.
Acc
essed:
June
20,
2020
[Onli
ne]
.
Avail
ab
le:
htt
ps://
ww
w.bri
t
anni
c
a.com/anim
al
/c
ri
cke
t
-
inse
ct
[20]
E.
A.
Shum
anov,
“
The
Cic
ad
a,
Cic
adetta
m
ont
ana
Scop.
,
as
a
Forest
Pest.
,”
Doklad
y
Akad
e
m
ii
nauk
SS
SR
,
vol.
72
,
no
.
6
,
pp
.
1127
-
1130
,
19
50
.
[21]
H.
W
ang
et
al
.
,
“
A
m
et
hod
of
high
throughput
m
onit
oring
cro
p
ph
y
sio
log
y
us
ing
chl
oroph
y
l
l
fluore
sce
n
ce
an
d
m
ult
ispec
tr
al
imaging
,”
Fronti
er
s in
P
lant
Scienc
e
,
vo
l.
9,
2018
,
d
oi:
10
.
3389/fpl
s
.
2018.
00407
.
[22]
J.
G.
A.
B
arb
ed
o,
“
Digit
a
l
imag
e
proc
essing
tec
hnique
s
for
de
tecti
ng
,
quan
ti
f
y
in
g
and
c
la
ss
if
y
in
g
pla
nt
disea
ses
,”
SpringerPl
us
,
vo
l.
2
,
no
.
1
,
pp
.
1
-
12
,
2013
,
d
oi
:
10
.
1186/2193
-
180
1
-
2
-
660
.
[23]
F.
Spinel
li,
A.
Cel
li
n
i,
L
.
Marc
het
ti,
K.
M.
Na
gesh,
and
C.
Pi
ovene
,
“
Emiss
ion
and
func
ti
on
of
vola
tile
orga
nic
compounds
in
r
esponse
to
abi
ot
ic
stress
,”
in
Ab
iot
ic
stress
in
plant
s
–
mec
hanisms
and
adaptat
ions
,
pp.
367
-
394
,
2011,
doi
:
10
.
57
72/24155
.
[24]
F.
Bril
l
i,
F.
Lor
e
to,
and
I.
Ba
cc
e
l
li
,
“
Exp
loi
t
ing
pl
ant
vol
at
i
le
orga
nic
compounds
(VO
Cs)
in
agr
ic
u
lt
ure
to
improve
sus
ta
ina
bl
e
def
ense
strat
eg
ie
s
and
produc
ti
v
ity
o
f
cro
ps
,”
Fr
onti
ers
in
Pl
ant
Scienc
e
,
vol.
10
,
2019,
doi:
10
.
3389/fpls
.
2019.
00264
.
[25]
M.
Saifuz
za
m
a
n,
N.
N.
Moon
and
F.
N.
Nu
r,
"IoT
base
d
s
tre
e
t
li
ghti
ng
a
nd
tra
ffic
m
anagem
ent
s
y
stem,
"
2017
IEEE
Re
gion
10
Hum
anit
arian
Technol
ogy
C
onfe
renc
e
(
R1
0
-
HT
C
)
,
2017,
pp.
121
-
12
4,
doi:
10
.
1109/R1
0
-
HTC.
2017.
82
88921
.
Evaluation Warning : The document was created with Spire.PDF for Python.