I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
24
,
p
p
.
7
1
9
8
~
7
2
1
0
I
SS
N:
2088
-
8
7
0
8
,
DOI
:
1
0
.
1
1
5
9
1
/ijece.
v
14
i
6
.
pp
7
1
9
8
-
7
2
1
0
7198
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
A
deep
learning
-
b
a
sed surv
eilla
nce
sy
stem f
o
r en
ha
ncing
p
ublic
sa
fet
y
t
hro
ug
h
i
nt
ernet
o
f
thi
ng
s
a
nd dig
ital te
chno
lo
g
y
using
Ra
spberry
P
i
Sh
re
edev
i K
a
re
pp
a
Sa
na
pa
n
na
v
a
r,
Cha
y
a
dev
i My
s
o
re
L
a
k
s
hm
a
na
g
o
wda
,
G
ee
t
ha
Su
nd
a
ra
ra
j
a
n
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
r
i
n
g
,
B
N
M
I
n
st
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
5
,
2
0
2
4
R
ev
is
ed
J
u
l 3
1
,
2
0
2
4
Acc
ep
ted
Au
g
6
,
2
0
2
4
In
p
u
b
li
c
sp
a
c
e
s,
in
d
i
v
i
d
u
a
ls
e
n
c
o
u
n
ter
c
h
a
ll
e
n
g
e
s
d
u
e
to
th
e
p
re
v
a
len
c
e
o
f
m
a
li
c
io
u
s
a
c
ti
v
it
ies
li
k
e
t
h
e
ft
a
n
d
k
id
n
a
p
p
i
n
g
.
As
th
e
in
ter
n
e
t
o
f
th
in
g
s
(Io
T)
a
n
d
d
ig
i
tal
tec
h
n
o
l
o
g
y
c
o
n
ti
n
u
e
to
e
x
p
a
n
d
ra
p
id
l
y
,
e
ffo
rts
to
c
re
a
te
sa
fe
e
n
v
iro
n
m
e
n
ts
a
re
b
e
c
o
m
in
g
in
c
re
a
sin
g
ly
so
p
h
isti
c
a
ted
.
T
o
a
d
d
r
e
ss
th
e
se
se
c
u
rit
y
c
o
n
c
e
rn
s,
a
p
r
o
p
o
se
d
s
o
lu
ti
o
n
in
v
o
l
v
e
s
th
e
u
ti
li
z
a
ti
o
n
o
f
v
i
d
e
o
-
c
a
p
tu
rin
g
tec
h
n
o
lo
g
y
with
t
h
e
h
e
lp
o
f
a
Ra
sp
b
e
rry
P
i
we
b
c
a
m
e
ra
.
Vid
e
o
s
o
f
th
e
su
rro
u
n
d
in
g
s
a
re
re
c
o
rd
e
d
,
a
d
ig
it
a
l
sig
n
a
tu
re
a
lg
o
rit
h
m
is
a
p
p
li
e
d
t
o
p
ro
tec
t
t
h
e
v
i
d
e
o
s,
a
n
d
th
e
y
a
re
th
e
n
tran
sm
it
te
d
to
a
u
th
o
rize
d
i
n
d
iv
i
d
u
a
ls
wh
o
u
se
th
e
m
f
o
r
f
o
re
n
sic
a
n
a
l
y
si
s.
Th
is
p
r
o
c
e
ss
a
ll
o
ws
fo
r
th
e
id
e
n
ti
fica
ti
o
n
a
n
d
i
n
v
e
sti
g
a
ti
o
n
o
f
a
n
y
su
sp
ici
o
u
s
o
r
c
rimin
a
l
a
c
ti
v
it
ies
.
Th
e
c
a
p
t
u
re
d
v
id
e
o
d
a
ta
is
c
o
m
p
a
re
d
with
a
sta
n
d
a
rd
d
a
tas
e
t
u
sin
g
a
d
e
e
p
lea
rn
in
g
p
r
o
c
e
ss
.
By
a
n
a
ly
z
in
g
th
e
c
o
n
ten
t
o
f
t
h
e
v
i
d
e
o
s
a
n
d
i
d
e
n
ti
f
y
in
g
t
h
e
p
o
te
n
t
ial
th
re
a
t
o
b
jec
ts,
we
c
a
n
a
ll
o
w
fo
r
p
ro
m
p
t
in
ter
v
e
n
ti
o
n
o
r
f
u
rth
e
r
i
n
v
e
sti
g
a
ti
o
n
b
y
re
lev
a
n
t
a
u
t
h
o
rit
ies
.
K
ey
w
o
r
d
s
:
C
lo
u
d
s
to
r
ag
e
E
llip
tic
cu
r
v
e
d
ig
ital sig
n
atu
r
e
a
lg
o
r
ith
m
I
n
ter
n
et
o
f
th
in
g
s
Pu
b
lic
s
p
ac
es
VGG1
6
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Sh
r
ee
d
ev
i K
ar
ep
p
a
San
ap
a
n
n
a
v
ar
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
B
NM
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
27
th
cr
o
s
s
,
1
2
th
Ma
in
,
B
SK I
I
s
tag
e,
B
en
g
alu
r
u
5
6
0
0
7
0
,
I
n
d
ia
E
m
ail: sh
r
ee
d
ev
is
r
@
b
n
m
it.in
1.
I
NT
RO
D
UCT
I
O
N
T
h
is
wo
r
k
u
tili
ze
s
th
e
in
ter
n
et
o
f
th
in
g
s
(
I
o
T
)
,
wh
ich
d
en
o
tes
to
a
n
etwo
r
k
o
f
in
ter
co
n
n
ec
t
ed
d
ev
ices
ca
p
ab
le
o
f
co
m
m
u
n
icatin
g
an
d
ex
ch
a
n
g
in
g
d
ata
ac
r
o
s
s
th
e
I
n
ter
n
et
with
o
u
t
h
u
m
a
n
in
ter
v
e
n
tio
n
.
T
h
e
r
an
g
e
o
f
s
u
ch
d
ev
ices
ca
n
ex
ten
d
f
r
o
m
s
im
p
le
s
en
s
o
r
s
an
d
ac
tu
ato
r
s
to
co
m
p
lex
in
d
u
s
tr
ial
m
ac
h
in
er
y
an
d
c
o
n
s
u
m
er
elec
tr
o
n
ics.
C
lo
u
d
s
to
r
a
g
e
is
a
s
er
v
ice
th
at
allo
ws
y
o
u
t
o
s
to
r
e
d
ata
s
ec
u
r
ely
o
n
r
em
o
te
s
er
v
er
s
ac
ce
s
s
ed
v
ia
th
e
in
ter
n
et,
r
ath
er
th
an
s
to
r
in
g
it
o
n
y
o
u
r
lo
ca
l
h
a
r
d
d
r
iv
e
o
r
o
th
er
p
h
y
s
ical
s
to
r
ag
e
d
ev
i
ce
s
.
W
h
en
u
tili
zin
g
clo
u
d
s
to
r
ag
e,
y
o
u
r
d
ata
is
s
to
r
ed
o
n
s
er
v
er
s
m
ain
tain
ed
b
y
a
th
ir
d
-
p
ar
ty
p
r
o
v
id
e
r
,
wh
ich
ar
e
o
f
ten
lo
ca
ted
in
d
ata
ce
n
ter
s
ar
o
u
n
d
th
e
wo
r
l
d
.
T
h
e
ellip
tic
cu
r
v
e
d
ig
ital
s
ig
n
atu
r
e
alg
o
r
ith
m
(
E
C
DSA)
is
a
cr
y
p
to
g
r
ap
h
ic
ap
p
r
o
ac
h
em
p
l
o
y
ed
t
o
cr
ea
te
d
ig
ital
s
ig
n
atu
r
es,
r
ely
in
g
o
n
e
llip
tic
cu
r
v
e
cr
y
p
to
g
r
ap
h
y
.
I
t
i
s
a
v
ar
iatio
n
o
f
th
e
d
ig
ital
s
ig
n
atu
r
e
alg
o
r
ith
m
(
DSA)
s
tan
d
ar
d
ized
b
y
NI
ST.
E
C
DSA
o
f
f
er
s
a
h
ig
h
lev
el
o
f
s
ec
u
r
ity
with
r
elativ
ely
s
m
aller
k
ey
s
izes
co
m
p
ar
ed
t
o
o
th
er
tr
ad
itio
n
al
d
ig
ital
s
ig
n
atu
r
e
s
ch
em
es.
T
h
e
co
r
e
o
f
E
C
DSA
r
elies o
n
th
e
p
r
o
p
er
ties
o
f
ellip
tic
cu
r
v
es o
v
er
f
in
ite
f
ield
s
.
A
n
ellip
tic
cu
r
v
e
is
d
ef
i
n
ed
b
y
,
2
=
3
+
+
w
h
e
r
e
a
n
d
a
r
e
c
o
n
s
t
a
n
t
s
,
a
n
d
t
h
e
c
u
r
v
e
i
s
d
e
f
i
n
e
d
w
i
t
h
i
n
a
f
i
n
i
t
e
f
i
e
l
d
.
E
C
D
SA
f
u
n
c
t
i
o
n
s
w
i
t
h
i
n
s
p
e
ci
f
ic
e
l
l
i
p
ti
c
c
u
r
v
e
d
o
m
a
i
n
p
a
r
a
m
e
te
r
s
,
w
h
i
c
h
e
n
c
o
m
p
a
s
s
t
h
e
c
u
r
v
e
e
q
u
a
t
i
o
n
,
b
as
e
p
o
i
n
t
,
o
r
d
e
r
,
a
n
d
o
t
h
e
r
r
e
l
at
e
d
p
a
r
a
m
e
t
e
r
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
u
r
ve
illa
n
ce
s
ystem
fo
r
en
h
a
n
cin
g
p
u
b
lic
…
(
S
h
r
ee
d
ev
i Ka
r
ep
p
a
S
a
n
a
p
a
n
n
a
v
a
r
)
7199
VGG1
6
is
a
d
ee
p
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
ar
ch
itectu
r
e
p
r
o
p
o
s
ed
b
y
th
e
v
is
u
al
g
eo
m
etr
y
g
r
o
u
p
(
VGG)
at
th
e
U
n
iv
er
s
ity
o
f
O
x
f
o
r
d
.
I
t
f
in
d
s
ex
ten
s
iv
e
a
p
p
li
ca
tio
n
in
im
ag
e
class
if
icatio
n
an
d
r
elate
d
task
s
in
co
m
p
u
ter
v
is
io
n
.
T
h
e
“
16
”
i
n
VGG1
6
d
en
o
tes
th
e
n
u
m
b
e
r
o
f
weig
h
t
lay
e
r
s
in
th
e
n
etw
o
r
k
,
e
x
clu
d
in
g
th
e
p
o
o
lin
g
a
n
d
ac
tiv
atio
n
lay
er
s
.
I
t
i
s
ch
ar
ac
ter
ized
b
y
its
s
im
p
licity
,
u
s
in
g
s
m
all
3
×
3
f
ilter
s
an
d
d
ee
p
er
n
etwo
r
k
s
.
T
h
e
ar
ch
itectu
r
e
h
a
s
n
o
co
m
p
lex
elem
en
ts
lik
e
r
esid
u
al
co
n
n
ec
tio
n
s
o
r
in
ce
p
tio
n
m
o
d
u
les,
m
ak
in
g
it
s
tr
aig
h
tf
o
r
war
d
to
u
n
d
er
s
tan
d
an
d
im
p
lem
e
n
t.
I
t
i
s
co
m
p
u
tatio
n
ally
in
ten
s
iv
e
co
m
p
a
r
ed
to
m
o
r
e
m
o
d
er
n
ar
ch
itectu
r
es
lik
e
r
esid
u
al
n
etwo
r
k
(
R
esNet)
an
d
I
n
ce
p
tio
n
d
u
e
to
its
d
ee
p
er
d
esig
n
.
T
h
is
ar
ticle
p
r
o
p
o
s
es
a
s
y
s
tem
th
at
estab
lis
h
es a
co
n
f
id
en
t a
n
d
s
ec
u
r
e
p
u
b
lic
s
p
ac
e
with
in
a
s
m
ar
t c
ity
.
Ou
r
r
esear
ch
aim
s
t
o
c
r
ea
te
an
in
tellig
en
t
s
u
r
v
eillan
ce
s
ec
u
r
ity
s
y
s
tem
ca
p
a
b
le
o
f
d
etec
tin
g
wea
p
o
n
s
,
s
p
ec
if
ically
g
u
n
s
o
r
k
n
iv
es.
I
n
p
u
r
s
u
it
o
f
th
is
o
b
jectiv
e,
w
e
u
tili
ze
d
d
i
v
er
s
e
co
m
p
u
ter
v
i
s
io
n
tech
n
iq
u
es
an
d
d
ee
p
lear
n
in
g
al
g
o
r
ith
m
s
to
id
en
tify
wea
p
o
n
s
f
r
o
m
ca
p
tu
r
ed
im
ag
es.
R
ec
en
t
a
d
v
an
ce
m
en
ts
in
m
ac
h
i
n
e
-
lear
n
in
g
an
d
d
ee
p
lear
n
i
n
g
,
p
ar
ticu
lar
ly
with
m
o
d
els
lik
e
VGG1
6
,
h
av
e
s
h
o
wn
s
ig
n
if
ican
t
p
r
o
g
r
ess
in
d
etec
tin
g
o
b
jects
an
d
r
ec
o
g
n
i
tio
n
,
esp
ec
ially
i
n
im
ag
es.
Ob
ject
id
en
tific
atio
n
an
d
class
if
icatio
n
ar
e
cr
u
cial
in
itial
s
tep
s
in
an
y
v
id
e
o
s
u
r
v
eillan
ce
s
y
s
tem
,
as
th
ey
p
a
v
e
th
e
way
f
o
r
f
u
r
th
e
r
o
b
ject
-
tr
ac
k
in
g
task
s
.
T
o
ac
co
m
p
lis
h
th
is
,
we
tr
ain
ed
a
class
if
ier
m
o
d
el
u
s
in
g
th
e
C
OC
O,
i.e
.
,
“
C
o
m
m
o
n
o
b
jects in
co
n
te
xt
”
d
ataset.
T
h
is
p
ap
er
aim
s
t
o
cr
ea
te
a
s
y
s
tem
ca
p
ab
le
o
f
s
wif
tly
d
e
tectin
g
g
u
n
s
an
d
k
n
iv
es
wh
il
e
u
tili
zin
g
m
in
im
al
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
.
W
ith
th
e
r
ap
id
ad
v
an
ce
m
en
t
o
f
tech
n
o
lo
g
y
,
it
h
as
b
ec
o
m
e
in
cr
ea
s
in
g
ly
cr
u
cial
to
d
etec
t
an
y
p
o
te
n
tial
th
r
ea
ts
in
p
u
b
lic
ar
ea
s
.
T
h
er
e
f
o
r
e,
o
u
r
p
r
o
p
o
s
ed
s
y
s
tem
h
as
th
e
p
o
ten
tial
to
b
e
ef
f
ec
tiv
ely
in
teg
r
ated
in
to
th
e
s
u
r
v
eillan
ce
s
y
s
tem
.
T
h
e
p
r
o
p
o
s
ed
s
ec
u
r
ity
m
ea
s
u
r
es
to
p
r
o
m
p
tly
id
en
tif
y
f
ir
ea
r
m
s
o
r
e
d
g
ed
wea
p
o
n
s
ar
e
s
h
o
wn
in
Fig
u
r
e
1
.
Fig
u
r
e
1
.
W
ea
p
o
n
s
d
etec
tio
n
t
y
p
es
2.
L
I
T
E
R
AT
U
RE
SU
RVE
Y
Par
k
in
g
s
ca
r
city
in
u
r
b
a
n
ar
ea
s
,
in
clu
d
in
g
u
n
iv
er
s
ity
ca
m
p
u
s
es,
n
ec
ess
itate
s
a
s
m
ar
t
p
a
r
k
i
n
g
s
y
s
tem
.
J
ab
b
ar
et
a
l.
[
1
]
in
tr
o
d
u
ce
s
an
I
o
T
-
b
ased
R
asp
b
er
r
y
Pi
p
ar
k
in
g
m
a
n
ag
em
e
n
t
s
y
s
tem
(
I
o
T
-
PiP
MS)
th
at
ef
f
icien
tly
m
atch
es
d
r
iv
er
s
with
v
ac
an
t
s
p
o
ts
,
r
ed
u
ce
s
co
n
g
esti
o
n
an
d
s
av
es
co
s
ts
.
T
h
e
s
y
s
tem
m
o
n
ito
r
s
p
ar
k
in
g
o
cc
u
p
an
c
y
,
g
u
id
es
d
r
iv
er
s
v
ia
a
s
m
ar
tp
h
o
n
e
ap
p
,
an
d
u
p
d
ates
d
ata
i
n
r
ea
l
tim
e.
I
t
e
n
h
an
ce
s
ac
ce
s
s
ib
ilit
y
an
d
u
s
er
ex
p
er
ie
n
ce
,
with
f
o
r
th
c
o
m
in
g
in
te
n
tio
n
s
f
o
r
ex
p
a
n
s
io
n
an
d
im
p
r
o
v
e
d
p
r
iv
ac
y
m
ea
s
u
r
es.
Stu
d
ies
[
2
]
p
r
esen
ts
a
n
o
v
el
u
n
if
ied
m
eth
o
d
f
o
r
au
to
m
atica
lly
d
etec
tin
g
v
eh
icles
an
d
licen
s
e
p
lates
in
u
r
b
an
s
u
r
v
eillan
ce
s
y
s
tem
s
.
Ou
r
ap
p
r
o
ac
h
id
en
tifie
s
h
ig
h
-
e
n
er
g
y
f
r
eq
u
en
cy
ar
ea
s
in
d
ig
ital c
am
er
a
im
ag
es,
r
ed
u
cin
g
d
ata
v
o
l
u
m
e.
W
e
in
t
r
o
d
u
ce
a
n
o
v
el
f
ilter
f
o
r
th
is
p
u
r
p
o
s
e
an
d
s
h
o
w
its
ef
f
ec
tiv
en
ess
in
I
o
T
a
n
d
s
m
ar
t
city
ap
p
licatio
n
s
.
T
ec
h
n
o
lo
g
y
h
as r
ev
o
lu
tio
n
ize
d
o
u
r
wo
r
ld
,
i
n
cl
u
d
in
g
h
o
m
e
s
ec
u
r
ity
[
3
]
.
W
e
p
r
o
p
o
s
e
a
R
asp
b
er
r
y
Pi
-
b
ased
h
o
m
e
s
ec
u
r
ity
s
y
s
te
m
u
s
in
g
h
ar
d
war
e
lik
e
R
asp
b
er
r
y
Pi,
a
ca
m
er
a,
a
t
o
u
ch
s
cr
ee
n
,
an
d
a
n
An
d
r
o
id
m
o
b
ile.
Ou
r
s
y
s
tem
,
a
s
m
ar
t
m
ir
r
o
r
,
s
er
v
es b
o
th
in
f
o
r
m
atio
n
al
an
d
s
ec
u
r
ity
p
u
r
p
o
s
es,
r
es
p
o
n
d
in
g
to
to
u
ch
an
d
m
o
b
ile
co
m
m
an
d
s
.
I
t
s
en
d
s
aler
t
m
ess
ag
es
with
clea
r
in
tr
u
d
er
p
h
o
to
s
to
th
e
o
wn
e
r
'
s
m
o
b
ile
u
p
o
n
d
etec
tio
n
.
T
r
ain
ed
with
y
o
u
o
n
ly
lo
o
k
o
n
ce
(
YOL
O)
an
d
Haa
r
tec
h
n
iq
u
es,
o
u
r
s
y
s
tem
ac
h
iev
e
s
9
6
%
ac
cu
r
ac
y
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
7
1
9
8
-
7
2
1
0
7200
d
etec
tin
g
h
u
m
an
in
tr
u
d
er
s
.
I
t
ca
n
p
o
ten
tially
a
u
th
en
ticate
u
s
er
s
b
y
co
m
p
ar
in
g
ca
p
tu
r
e
d
f
ac
es
with
s
to
r
ed
p
h
o
to
s
,
b
u
t th
is
p
r
esen
ts
ch
allen
g
es in
clea
r
im
ag
e
ca
p
tu
r
e
d
u
r
in
g
a
u
th
en
ticatio
n
.
As u
r
b
an
p
ar
k
in
g
ch
allen
g
es
[
4
]
g
r
o
w
with
in
c
r
ea
s
in
g
v
e
h
i
cle
n
u
m
b
e
r
s
,
th
is
ar
ticle
in
tr
o
d
u
ce
s
a
d
u
al
-
len
s
m
illi
m
eter
wav
e
(
MM
W
)
r
ad
ar
an
ten
n
a
f
o
r
I
o
T
-
b
ased
s
m
ar
t
p
ar
k
in
g
s
y
s
tem
s
.
I
t
f
ea
tu
r
es
a
f
lat
d
ielec
tr
ic
p
u
n
ch
len
s
to
b
o
o
s
t
tr
an
s
m
itti
n
g
an
ten
n
a
g
ain
a
n
d
a
d
ielec
tr
ic
r
o
d
le
n
s
f
o
r
wid
e
b
ea
m
wid
th
an
d
s
tab
le
p
er
f
o
r
m
an
ce
.
T
h
e
co
m
b
in
ed
d
u
al
-
len
s
d
esig
n
en
h
a
n
ce
s
r
ad
ar
ac
cu
r
ac
y
an
d
s
tab
ilit
y
at
2
4
GHz
.
T
r
an
s
m
itti
n
g
an
ten
n
a
g
ain
m
ea
s
u
r
es
1
5
.
8
d
B
i,
r
ec
eiv
in
g
an
ten
n
a
g
ain
7
.
9
d
B
i
,
with
a
3
d
B
b
ea
m
wid
th
o
f
6
5
d
eg
r
ee
s
.
T
h
e
an
ten
n
a
o
f
f
er
s
s
tab
le
p
er
f
o
r
m
a
n
ce
,
s
u
itab
le
f
o
r
MM
W
r
ad
ar
s
m
ar
t
p
ar
k
in
g
s
y
s
tem
s
,
an
d
is
co
s
t
-
e
f
f
ec
tiv
e
co
m
p
ar
ed
to
ar
r
a
y
-
b
as
ed
s
o
lu
tio
n
s
.
C
lo
u
d
co
m
p
u
ti
n
g
is
m
ain
s
tr
ea
m
b
u
t
with
u
r
b
an
izatio
n
,
t
h
er
e
i
s
m
o
r
e
s
u
r
v
eillan
ce
d
ata
[
5
]
.
E
d
g
e
co
m
p
u
tin
g
o
f
f
er
s
a
s
o
lu
tio
n
.
Ou
r
s
tu
d
y
in
v
esti
g
ates
u
s
in
g
it
f
o
r
s
m
ar
t
p
ar
k
in
g
s
u
r
v
eillan
ce
.
W
e
ac
h
iev
ed
o
v
er
9
5
%
ac
cu
r
ac
y
in
r
ea
l
-
wo
r
ld
te
s
ts
.
T
h
is
s
y
s
tem
co
u
ld
b
e
cr
u
c
ial
f
o
r
s
m
ar
t
cities
an
d
tr
an
s
p
o
r
tatio
n
.
Ou
r
p
ap
er
d
escr
ib
es
th
e
d
ev
elo
p
m
en
t
an
d
test
in
g
o
f
th
is
s
m
ar
t
p
ar
k
in
g
s
u
r
v
eillan
ce
s
y
s
tem
,
wh
ich
is
am
o
n
g
t
h
e
f
ir
s
t
to
u
s
e
ed
g
e
co
m
p
u
tin
g
in
r
ea
l
-
w
o
r
ld
p
ar
k
in
g
s
u
r
v
eillan
ce
.
W
e
ac
h
iev
ed
9
5
.
6
%
ac
cu
r
ac
y
in
d
iv
er
s
e
co
n
d
itio
n
s
,
o
f
f
er
in
g
ad
v
an
tag
es
o
v
er
ex
i
s
tin
g
s
y
s
tem
s
an
d
p
o
ten
tial
f
o
r
s
m
ar
t
cities
an
d
tr
a
n
s
p
o
r
t
atio
n
.
Ku
m
b
h
ar
[
6
]
in
tr
o
d
u
ce
s
a
n
I
o
T
-
b
ased
s
ec
u
r
ity
s
y
s
tem
f
o
r
r
esid
en
ce
s
u
t
ilizin
g
R
asp
b
er
r
y
Pi
3
,
Pi
ca
m
er
a,
a
n
d
p
ass
iv
e
in
f
r
ar
ed
(
PIR
)
s
en
s
o
r
.
I
t
s
en
d
s
em
ail
aler
ts
to
h
o
m
eo
wn
er
s
an
d
m
em
b
e
r
s
u
p
o
n
d
etec
tin
g
in
tr
u
s
io
n
,
an
d
ca
n
also
ac
tiv
ate
a
s
ec
u
r
ity
alar
m
.
Peo
p
le
h
a
v
e
th
e
ab
ilit
y
to
m
o
n
ito
r
r
e
m
o
tely
t
h
eir
r
esid
en
ce
s
an
d
r
ec
ei
v
e
aler
ts
o
n
th
eir
d
e
v
ices,
p
r
e
v
en
tin
g
u
n
au
th
o
r
ize
d
ac
ce
s
s
.
T
h
e
in
ter
n
et
o
f
th
in
g
s
(
I
o
T
)
m
e
r
g
es
d
e
v
ices
in
to
n
etwo
r
k
s
f
o
r
ad
v
a
n
ce
d
s
er
v
ices,
b
u
t
p
r
i
v
ac
y
an
d
s
ec
u
r
ity
a
r
e
cr
itical
[
7
]
.
T
h
i
s
p
a
p
e
r
e
x
p
l
o
r
e
s
I
o
T
a
t
ta
c
k
m
o
d
e
l
s
a
n
d
M
L
-
b
a
s
e
d
s
e
c
u
r
it
y
s
o
l
u
ti
o
n
s
,
f
o
c
u
s
i
n
g
o
n
a
u
t
h
e
n
t
i
c
a
t
i
o
n
,
a
c
c
es
s
co
n
t
r
o
l
,
m
a
l
w
a
r
e
d
e
t
e
ct
i
o
n
,
a
n
d
s
e
c
u
r
e
o
f
f
l
o
a
d
i
n
g
.
C
h
a
l
le
n
g
e
s
r
e
m
a
i
n
i
n
i
m
p
l
e
m
e
n
t
i
n
g
t
h
e
s
e
t
e
c
h
n
i
q
u
es
i
n
r
e
a
l
-
w
o
r
l
d
I
o
T
s
y
s
t
e
m
s
.
K
i
e
u
-
Do
-
N
g
u
y
e
n
e
t
a
l
.
[
8
]
i
n
t
r
o
d
u
c
e
s
a
m
u
l
t
i
-
f
u
n
c
ti
o
n
a
l
e
l
l
i
p
t
i
c
c
u
r
v
e
c
r
y
p
t
o
g
r
a
p
h
y
(
E
C
C
)
h
a
r
d
w
a
r
e
d
es
i
g
n
s
u
p
p
o
r
t
i
n
g
E
C
D
S
A
a
n
d
E
d
w
a
r
d
s
-
c
u
r
v
e
d
i
g
it
a
l
s
i
g
n
a
t
u
r
e
a
l
g
o
r
i
t
h
m
(
E
d
D
S
A
)
a
l
g
o
r
i
t
h
m
s
.
I
t
i
s
e
f
f
i
ci
e
n
t
,
l
o
w
-
c
o
s
t
,
an
d
d
o
e
s
not
r
el
y
o
n
d
i
g
i
t
a
l
s
i
g
n
a
l
p
r
o
c
es
s
o
r
(
DS
P
)
.
T
h
e
c
o
r
e
r
u
n
s
u
p
t
o
1
1
2
.
2
M
H
z
w
i
t
h
V
i
r
t
e
x
-
7
d
e
v
i
c
es
,
u
s
i
n
g
o
n
l
y
1
0
,
2
5
9
s
l
i
c
e
s
.
I
t
a
c
c
o
m
m
o
d
a
t
e
s
m
u
lt
i
p
l
e
E
C
C
a
l
g
o
r
i
t
h
m
s
(
E
d
2
5
5
1
9
,
E
C
DS
A
w
it
h
N
I
S
T
P
-
2
5
6
,
P
-
3
8
4
,
a
n
d
P
-
5
2
1
)
i
n
o
n
e
d
e
s
i
g
n
,
e
n
a
b
l
i
n
g
it
s
c
o
m
p
a
ti
b
i
lit
y
w
i
t
h
d
i
v
e
r
s
e
h
a
r
d
w
a
r
e
p
l
at
f
o
r
m
s
w
i
t
h
o
u
t
n
e
ce
s
s
i
ta
t
i
n
g
r
e
d
esi
g
n
.
Pu
th
iy
id
am
et
a
l.
[
9
]
s
u
g
g
es
ts
a
tim
estam
p
-
b
ased
ap
p
r
o
a
ch
to
im
p
r
o
v
e
E
C
DSA
f
o
r
I
o
T
d
ev
ice
au
th
en
ticatio
n
.
I
t
p
r
ev
en
ts
s
ig
n
atu
r
e
r
e
u
s
e
an
d
f
ak
e
s
ig
n
atu
r
e
g
en
er
atio
n
b
y
g
e
n
er
atin
g
u
n
i
q
u
e
s
ig
n
atu
r
es
ea
ch
tim
e.
T
h
is
en
h
an
ce
s
s
ec
u
r
ity
a
n
d
ef
f
icien
c
y
,
o
u
t
p
er
f
o
r
m
in
g
m
o
s
t
E
C
DSA
v
ar
ian
ts
.
Ou
r
ex
p
er
im
en
ts
co
n
f
ir
m
its
ef
f
ec
tiv
en
ess
with
m
in
i
m
al
o
v
er
h
ea
d
.
Fu
tu
r
e
wo
r
k
in
clu
d
es
im
p
r
o
v
in
g
ex
ec
u
t
io
n
ef
f
icien
cy
a
n
d
ad
d
r
ess
in
g
q
u
an
tu
m
co
m
p
u
ti
n
g
s
ec
u
r
ity
.
He
et
a
l.
[
1
0
]
p
r
o
p
o
s
es
a
lo
w
-
laten
cy
E
C
DSA
ar
ch
itectu
r
e,
s
ig
n
if
ican
tly
r
ed
u
cin
g
d
o
u
b
le
p
o
in
t
m
u
ltip
licatio
n
(
DPM)
ca
lcu
latio
n
to
f
iv
e
clo
ck
cy
cles
p
er
lo
o
p
.
Ou
r
d
esig
n
o
u
tp
er
f
o
r
m
s
ex
is
tin
g
o
n
es,
ac
h
iev
in
g
DPM
tim
es
o
f
3
.
5
8
4
,
5
.
6
5
6
,
an
d
7
.
4
5
3
µ
s
o
v
er
GF(
2
1
6
3
)
,
GF(
2
2
3
3
)
,
an
d
GF(
2
2
8
3
)
,
r
esp
ec
tiv
ely
,
with
im
p
r
o
v
ed
th
r
o
u
g
h
p
u
t e
f
f
icien
cy
.
Sm
ar
t
h
o
m
es
im
p
r
o
v
e
liv
in
g
s
tan
d
ar
d
s
b
u
t
f
a
ce
d
ata
s
e
cu
r
ity
ch
allen
g
es
[
1
1
]
.
W
e
p
r
o
p
o
s
e
a
b
lo
ck
ch
ain
-
b
ased
m
u
lti
-
clo
u
d
s
to
r
ag
e
m
o
d
el
an
d
a
n
ef
f
i
cien
t
id
en
tity
-
b
ased
p
r
o
x
y
a
g
g
r
eg
ate
s
ig
n
at
u
r
e
(
I
B
PAS)
s
ch
em
e
to
en
h
an
c
e
s
ec
u
r
ity
a
n
d
c
o
n
s
er
v
e
r
e
s
o
u
r
ce
s
.
Ou
r
ex
p
e
r
im
en
ts
d
em
o
n
s
tr
ate
n
o
tab
le
en
h
an
ce
m
e
n
ts
in
co
m
m
u
n
icatio
n
ex
p
en
s
es
an
d
s
to
r
a
g
e
ef
f
ec
tiv
en
ess
wh
en
ju
x
tap
o
s
ed
with
co
n
v
en
tio
n
al
m
eth
o
d
s
.
T
h
is
m
eth
o
d
o
lo
g
y
g
u
ar
an
tees
th
e
in
teg
r
ity
an
d
d
e
p
en
d
ab
ilit
y
o
f
d
ata
in
s
m
ar
t
h
o
m
e
en
v
ir
o
n
m
en
ts
.
R
ajash
r
ee
et
a
l.
[
1
2
]
in
tr
o
d
u
ce
s
a
R
asp
b
er
r
y
Pi
-
b
ased
s
ec
u
r
it
y
s
y
s
tem
f
o
r
en
tr
a
n
ce
v
is
ito
r
v
alid
atio
n
,
u
tili
zin
g
E
C
DSA
w
ith
Wi
-
Fi
.
I
t
ca
p
tu
r
es,
tim
estam
p
s
,
an
d
au
th
en
ticates
v
is
ito
r
im
ag
es,
s
en
d
in
g
th
em
f
o
r
v
er
if
icatio
n
v
ia
em
ail
o
n
W
in
d
o
ws
o
r
L
i
n
u
x
.
C
o
m
p
o
n
en
ts
in
clu
d
e
R
asp
b
er
r
y
Pi
3
,
R
asp
b
er
r
y
Pi
C
am
er
a,
MI
R
AC
L
s
o
f
twar
e
lib
r
ar
y
,
a
tactile
s
wit
ch
,
an
d
a
b
r
ea
d
b
o
a
r
d
.
T
h
e
l
a
t
e
s
t
t
r
e
n
d
s
l
i
k
e
I
o
T
,
s
m
a
r
t
c
i
t
i
es
,
a
n
d
d
i
g
i
t
a
l
t
r
a
n
s
f
o
r
m
a
t
io
n
a
r
e
f
u
e
l
i
n
g
h
u
g
e
d
a
t
a
g
r
o
w
t
h
,
m
a
k
i
n
g
c
l
o
u
d
s
t
o
r
a
g
e
c
r
u
c
i
a
l
[
1
3
]
.
Y
e
t
,
s
e
c
u
r
i
t
y
c
o
n
c
e
r
n
s
p
e
r
s
is
t
.
O
u
r
p
a
p
e
r
r
e
v
i
ew
s
c
l
o
u
d
s
t
o
r
a
g
e
s
e
c
u
r
it
y
c
o
m
p
r
e
h
e
n
s
i
v
e
l
y
,
c
o
v
e
r
i
n
g
c
h
a
l
l
e
n
g
e
s
,
e
n
c
r
y
p
t
i
o
n
m
e
t
h
o
d
s
,
a
n
d
o
p
e
n
r
e
s
e
a
r
c
h
t
o
p
i
cs
.
W
e
ed
s
a
r
e
a
s
u
b
s
ta
n
t
i
al
h
u
r
d
l
e
f
o
r
c
o
r
n
p
r
o
d
u
c
t
i
o
n
.
Y
an
g
e
t
a
l
.
[
1
4
]
p
r
o
p
o
s
e
d
SE
-
V
GG
1
6
,
a
d
e
e
p
l
ea
r
n
i
n
g
m
o
d
e
l
,
t
o
a
c
c
u
r
a
t
e
l
y
i
d
e
n
t
i
f
y
w
e
e
d
s
i
n
c
o
r
n
f
i
e
l
d
s
.
S
E
-
V
GG1
6
i
m
p
r
o
v
e
s
o
n
V
G
G
1
6
b
y
i
n
c
o
r
p
o
r
a
t
i
n
g
t
h
e
S
E
a
t
t
e
n
ti
o
n
m
e
c
h
a
n
i
s
m
,
r
e
d
u
c
i
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
k
e
r
n
e
l
s
i
ze
,
a
n
d
u
s
i
n
g
L
e
a
k
y
r
e
ct
i
f
i
e
d
l
i
n
e
a
r
u
n
i
t
(
L
e
a
k
y
R
e
L
U
)
a
ct
i
v
a
ti
o
n
f
u
n
c
t
i
o
n
s
.
I
t
a
c
h
i
e
v
e
s
9
9
.
6
7
%
a
c
c
u
r
a
c
y
,
o
u
t
p
e
r
f
o
r
m
in
g
o
t
h
e
r
m
o
d
e
l
s
.
T
h
e
s
t
u
d
ie
s
[
1
5
]
,
[
1
6
]
a
s
s
es
s
e
d
T
i
m
o
h
o
l
e
a
f
e
x
t
r
a
c
t
(
T
L
E
)
a
s
a
c
o
r
r
o
s
i
o
n
i
n
h
i
b
i
t
o
r
f
o
r
3
0
4
s
t
a
in
l
e
s
s
s
t
e
el
(
3
0
4
S
S
)
i
n
0
.
5
M
H
2
S
O
4
.
T
L
E
,
e
f
f
e
c
t
i
v
e
a
t
0
–
6
g
/
L
,
r
e
d
u
c
e
d
c
o
r
r
o
s
i
o
n
b
y
9
9
.
3
7
%
a
t
3
g
/
L
.
M
i
x
e
d
a
d
s
o
r
p
t
i
o
n
w
a
s
i
d
e
n
t
i
f
ie
d
a
s
t
h
e
i
n
h
i
b
i
t
i
o
n
m
e
c
h
a
n
is
m
.
M
a
c
h
in
e
l
e
a
r
n
i
n
g
(
C
NN
-
V
G
G
1
6
)
p
r
e
d
i
c
t
e
d
c
o
r
r
o
s
i
o
n
w
i
t
h
9
6
%
a
c
c
u
r
a
c
y
.
P
h
e
n
o
l
i
n
T
L
E
p
l
a
y
e
d
a
n
e
s
s
e
n
t
i
a
l
f
u
n
c
ti
o
n
i
n
a
d
s
o
r
p
t
i
o
n
o
n
3
0
4
S
S
.
Q
u
a
n
t
u
m
c
h
e
m
i
s
t
r
y
im
p
r
o
v
e
d
c
o
m
p
r
e
h
e
n
s
i
o
n
o
f
t
h
e
a
d
s
o
r
p
t
i
o
n
p
r
o
c
e
s
s
.
A
d
v
a
n
c
em
e
n
t
s
i
n
s
e
n
s
o
r
t
e
c
h
a
n
d
b
r
o
a
d
b
a
n
d
n
e
t
w
o
r
k
s
a
r
e
d
r
i
v
i
n
g
t
h
e
e
m
e
r
g
e
n
c
e
o
f
t
h
e
i
n
t
e
r
n
e
t
o
f
e
v
e
r
y
t
h
i
n
g
(
I
o
E
)
a
n
d
m
o
b
i
l
e
I
o
T
[
1
7
]
.
H
o
w
e
v
e
r
,
m
o
b
i
l
e
I
o
T
n
e
tw
o
r
k
s
f
a
c
e
c
h
a
ll
e
n
g
e
s
f
r
o
m
c
o
m
p
l
e
x
c
o
m
m
u
n
i
c
a
t
i
o
n
e
n
v
i
r
o
n
m
e
n
t
s
a
n
d
e
n
e
r
g
y
-
i
n
t
e
n
s
i
v
e
d
a
t
a
p
r
o
c
e
s
s
i
n
g
.
I
n
r
e
s
p
o
n
s
e
t
o
t
h
i
s
,
w
e
s
u
g
g
e
s
t
an
i
m
p
r
o
v
e
d
C
N
N
-
b
a
s
e
d
a
l
g
o
r
i
t
h
m
f
o
r
p
r
e
d
i
c
t
i
n
g
o
u
t
a
g
e
p
r
o
b
a
b
i
l
i
t
y
(
O
P
)
i
n
m
o
b
i
l
e
I
o
T
n
e
t
w
o
r
k
s
,
i
n
t
e
g
r
a
t
i
n
g
t
r
a
n
s
m
i
t
a
n
t
e
n
n
a
s
el
e
c
ti
o
n
(
T
AS
)
a
n
d
c
o
o
p
e
r
a
t
i
v
e
s
c
h
e
m
e
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
u
r
ve
illa
n
ce
s
ystem
fo
r
en
h
a
n
cin
g
p
u
b
lic
…
(
S
h
r
ee
d
ev
i Ka
r
ep
p
a
S
a
n
a
p
a
n
n
a
v
a
r
)
7201
Ob
ject
d
etec
tio
n
is
cr
u
cial
in
a
u
to
n
o
m
o
u
s
d
r
iv
i
n
g
,
u
tili
zin
g
a
lg
o
r
ith
m
s
lik
e
YOL
Ov
3
o
n
d
a
tasets
lik
e
C
OC
O
to
id
en
tify
o
b
jects a
cc
u
r
ately
.
T
h
is
tech
n
o
l
o
g
y
is
ess
en
tial f
o
r
s
af
e
d
r
iv
in
g
,
as seen
in
in
n
o
v
atio
n
s
lik
e
T
esla's
s
elf
-
d
r
iv
in
g
ca
r
s
,
wh
i
ch
aim
to
r
ed
u
ce
p
o
llu
ti
o
n
a
n
d
tr
af
f
ic
[
1
8
]
.
O
b
ject
d
etec
t
io
n
en
s
u
r
es
th
ese
v
eh
icles
ca
n
ac
cu
r
ately
p
er
c
eiv
e
th
eir
s
u
r
r
o
u
n
d
i
n
g
s
,
d
etec
tin
g
v
e
h
icles,
p
ed
estrian
s
,
a
n
d
tr
a
f
f
ic
s
ig
n
s
to
p
r
ev
en
t
ac
ci
d
en
ts
an
d
e
n
s
u
r
e
s
af
e
tr
av
el.
T
h
is
p
ap
er
o
f
f
e
r
s
a
co
m
p
r
eh
e
n
s
iv
e
s
u
r
v
e
y
o
f
s
tate
-
of
-
th
e
-
ar
t
wea
p
o
n
d
etec
tio
n
m
eth
o
d
s
,
f
o
cu
s
in
g
o
n
s
p
ec
if
ic
c
h
allen
g
es
an
d
a
p
p
licatio
n
s
[
1
9
]
.
I
n
r
esp
o
n
s
e
t
o
in
cr
ea
s
in
g
cr
im
e,
Kay
a
et
a
l.
[
2
0
]
in
tr
o
d
u
ce
s
a
s
u
p
er
io
r
d
ee
p
lear
n
i
n
g
m
o
d
e
l,
ac
h
iev
in
g
9
8
.
4
0
%
ac
cu
r
ac
y
in
d
etec
tin
g
s
ev
en
wea
p
o
n
ty
p
es.
T
ested
o
n
d
iv
e
r
s
e
b
ac
k
g
r
o
u
n
d
s
,
it
en
h
an
ce
s
s
ec
u
r
ity
ef
f
icac
y
an
d
in
s
p
ir
es
f
u
r
th
er
r
esear
ch
in
r
ea
l
-
tim
e
m
o
n
ito
r
i
n
g
an
d
co
n
ce
aled
g
u
n
d
etec
tio
n
.
Nag
a
p
p
an
an
d
R
an
i
[
2
1
]
en
h
an
ce
s
s
ec
u
r
ity
th
r
o
u
g
h
r
ea
l
-
tim
e
wea
p
o
n
d
etec
tio
n
u
s
in
g
c
lo
s
ed
-
cir
cu
it
telev
is
io
n
(
C
C
T
V)
[
2
2
]
,
[
2
3
]
.
B
y
em
p
lo
y
in
g
a
d
v
an
ce
d
alg
o
r
ith
m
s
lik
e
YOL
O
v
5
,
we
ac
h
iev
ed
a
h
ig
h
F1
-
s
co
r
e
o
f
9
1
%,
e
n
s
u
r
in
g
s
af
ety
a
n
d
attr
ac
tin
g
in
v
e
s
tm
en
t.
Glo
b
al
g
u
n
v
io
len
ce
p
r
o
m
p
ts
ac
tio
n
[
2
4
]
YOL
O
v
3
s
y
s
tem
d
etec
ts
f
ir
ea
r
m
s
ef
f
icien
tly
,
e
n
h
an
cin
g
s
u
r
v
eillan
ce
.
Ou
tp
er
f
o
r
m
in
g
tr
ad
itio
n
al
m
et
h
o
d
s
,
it
o
f
f
er
s
co
s
t
-
ef
f
ec
tiv
e
s
o
lu
tio
n
s
,
r
e
v
o
lu
tio
n
izin
g
s
af
ety
th
r
o
u
g
h
r
o
b
o
tics
.
Ou
r
co
m
b
i
n
ed
[
2
5
]
Dee
p
E
AST
an
d
Ker
as
o
p
tical
ch
ar
ac
ter
r
ec
o
g
n
itio
n
(
OC
R
)
m
o
d
el
en
h
an
ce
s
tex
t
d
etec
tio
n
an
d
r
ec
o
g
n
itio
n
i
n
v
id
eo
s
.
W
h
ile
ef
f
ec
tiv
e
o
n
d
atasets
lik
e
I
C
DAR
2
0
1
5
,
im
p
r
o
v
em
en
ts
ar
e
n
ee
d
ed
,
p
ar
ticu
lar
ly
f
o
r
R
o
ad
T
ex
t
-
1
K.
Su
b
s
eq
u
en
t e
f
f
o
r
ts
aim
to
in
teg
r
ate
p
r
e
-
tr
ain
e
d
m
o
d
els to
en
h
an
ce
p
er
f
o
r
m
a
n
ce
ac
r
o
s
s
d
iv
er
s
e
co
n
d
itio
n
s
.
B
h
a
tti
et
a
l.
[
2
6
]
r
ef
e
r
s
to
test
in
g
v
ar
io
u
s
alg
o
r
ith
m
s
in
clu
d
in
g
VGG1
6
,
I
n
ce
p
tio
n
-
V3
,
an
d
I
n
ce
p
tio
n
-
R
esNetV2
.
C
h
ay
ad
ev
i
et
a
l.
[
2
7
]
cr
ea
t
ed
an
au
to
m
ate
d
s
ec
u
r
ity
s
y
s
tem
f
o
r
au
to
m
ated
teller
m
ac
h
in
es
(
AT
M
s
)
,
u
s
in
g
im
ag
e
p
r
o
ce
s
s
in
g
.
I
t
r
ec
o
g
n
izes
f
ac
es
with
V
io
la
-
J
o
n
es
an
d
d
etec
ts
wea
p
o
n
s
with
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
an
d
r
an
d
o
m
f
o
r
ests
.
T
en
s
o
r
Flo
w
d
r
iv
es
th
e
m
ac
h
in
e
lear
n
in
g
.
An
An
d
r
o
id
a
p
p
aler
ts
au
th
o
r
ities
,
an
d
th
er
e
i
s
a
s
p
ee
ch
aler
t f
o
r
th
e
v
is
u
ally
im
p
air
ed
.
3.
SYST
E
M
ARCH
I
T
E
CT
U
R
E
T
h
is
s
ec
tio
n
o
u
tlin
es
th
e
m
eth
o
d
o
lo
g
y
u
s
ed
,
r
esear
ch
m
ater
ials
em
p
lo
y
ed
,
s
y
s
tem
d
esig
n
,
an
d
f
ab
r
icatio
n
p
r
o
ce
s
s
.
T
h
e
id
ea
ex
p
lo
r
ed
in
t
h
is
p
ap
er
in
v
o
lv
es
co
m
b
in
in
g
I
o
T
with
t
h
e
id
ea
o
f
s
m
ar
t
p
u
b
lic
s
p
ac
es.
Ou
r
m
ain
f
o
cu
s
is
to
en
s
u
r
e
th
e
s
y
s
tem
o
p
er
ates
ef
f
ec
tiv
ely
in
b
o
th
t
h
e
p
r
o
to
t
y
p
e
an
d
r
ea
l
-
wo
r
ld
en
v
ir
o
n
m
en
ts
.
T
h
is
co
m
p
r
eh
e
n
s
iv
e
s
y
s
tem
en
s
u
r
es
b
o
th
th
e
s
ec
u
r
ity
o
f
th
e
ca
p
tu
r
e
d
f
o
o
tag
e
an
d
p
r
o
ac
tiv
e
m
ea
s
u
r
es in
r
esp
o
n
s
e
to
p
o
ten
t
ial
th
r
ea
ts
,
wh
ich
is
as sh
o
wn
in
Fig
u
r
e
2
.
Fig
u
r
e
2
.
R
asp
b
er
r
y
Pi
b
ased
s
m
ar
t p
u
b
lic
s
p
ac
es m
an
a
g
em
e
n
t sy
s
tem
ar
ch
itectu
r
e
3
.
1
.
Select
io
n o
f
co
m
po
nents
T
h
e
m
ater
ials
an
d
co
m
p
o
n
en
t
s
ar
e
s
elec
ted
b
ased
o
n
n
ee
d
s
an
d
co
m
p
atib
ilit
y
,
d
ete
r
m
in
ed
th
r
o
u
g
h
a
th
o
r
o
u
g
h
liter
atu
r
e
r
ev
iew.
Du
r
in
g
th
is
r
ev
iew,
we
ex
am
in
e
d
v
ar
io
u
s
d
esig
n
v
a
r
iatio
n
s
an
d
co
m
p
o
n
e
n
ts
u
s
ed
,
as
well
as
th
eir
f
u
n
ctio
n
s
in
s
m
ar
t
p
u
b
lic
p
r
em
is
es
s
y
s
tem
s
.
Har
d
war
e
a
n
d
s
o
f
twar
e
c
h
o
ic
es
ar
e
m
ad
e
d
u
r
in
g
th
e
co
m
p
o
n
en
t
s
elec
tio
n
p
h
a
s
e.
T
h
e
s
elec
ted
h
ar
d
war
e
c
o
m
p
o
n
en
ts
in
clu
d
e
R
asp
b
er
r
y
Pi
3
m
o
d
el
V1
.
3
,
R
asp
b
er
r
y
Pi
web
ca
m
er
a
with
USB
ca
b
le,
128
-
b
it
Mic
r
o
SD
ca
r
d
a
n
d
L
E
D
lig
h
ts
.
So
f
twar
e
to
o
ls
ch
o
s
en
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
7
1
9
8
-
7
2
1
0
7202
in
clu
d
e
th
e
R
asp
b
er
r
y
Pi
o
p
er
atin
g
s
y
s
tem
,
Py
th
o
n
,
m
u
lti
-
p
r
ec
is
io
n
in
teg
er
r
atio
n
al
ar
ith
m
etic
cr
y
p
to
lib
r
a
r
y
(
MI
R
AC
L
)
,
an
d
o
th
e
r
s
.
Af
ter
co
m
p
letin
g
th
e
c
o
m
p
o
n
en
t
s
elec
tio
n
,
th
e
s
y
s
tem
ar
c
h
itectu
r
e
is
estab
lis
h
ed
.
T
h
is
s
u
b
s
ec
tio
n
o
f
f
er
s
a
s
u
m
m
ar
y
o
f
th
e
ch
o
s
en
c
o
m
p
o
n
en
ts
.
T
h
e
in
ter
co
n
n
ec
tio
n
o
f
th
e
co
m
p
o
n
en
ts
is
d
ep
icted
in
Fig
u
r
e
3
.
Fig
u
r
e
3
.
C
o
n
n
ec
tio
n
b
etwe
en
R
asp
b
er
r
y
Pi an
d
Pi we
b
ca
m
er
a
3
.
1
.
1
.
Ra
s
pb
er
ry
P
i 3
M
o
del V
1
.
3
T
h
e
R
asp
b
er
r
y
Pi
3
m
o
d
el
V1
.
3
was
c
h
o
s
en
f
o
r
th
is
p
r
o
ject
p
r
im
a
r
ily
f
o
r
its
r
o
b
u
s
t
p
r
o
ce
s
s
in
g
ca
p
ab
ilit
ies
as
a
s
in
g
le
-
b
o
ar
d
em
b
ed
d
e
d
co
m
p
u
te
r
.
I
t
ex
ce
ls
in
co
n
n
ec
tin
g
to
th
e
in
ter
n
et
,
o
f
f
er
in
g
o
p
tio
n
s
v
ia
eith
er
an
E
th
er
n
et
p
o
r
t
o
r
wir
eless
ly
th
r
o
u
g
h
W
i
-
Fi
o
r
B
lu
eto
o
th
.
I
ts
ea
s
e
o
f
in
ter
n
et
co
n
n
ec
tiv
ity
s
tan
d
s
o
u
t
as
o
n
e
o
f
R
asp
b
er
r
y
Pi'
s
ad
v
an
tag
es
o
v
e
r
Ar
d
u
in
o
.
Ad
d
iti
o
n
ally
,
R
asp
b
er
r
y
Pi
s
u
p
p
o
r
ts
a
d
iv
er
s
e
r
an
g
e
o
f
p
r
o
g
r
a
m
m
in
g
lan
g
u
ag
es,
in
cl
u
d
in
g
Py
th
o
n
,
J
av
a,
C
,
C
++
,
Per
l,
R
u
b
y
,
an
d
B
ASI
C
.
I
n
co
n
tr
ast,
Ar
d
u
in
o
is
lim
ited
to
Ar
d
u
in
o
o
r
C
/C
++
.
I
t
h
as
B
r
o
ad
co
m
B
C
M2
8
3
7
B
0
,
C
o
r
tex
-
A5
3
(
AR
Mv
8
)
6
4
-
b
it
So
C
@
1
.
4
GHz
1
GB
L
PDDR
2
SDR
AM
.
T
h
e
Pi
web
ca
m
er
a
n
ee
d
s
s
u
b
s
tan
tial
p
r
o
ce
s
s
in
g
p
o
wer
to
c
ap
tu
r
e
a
n
d
p
r
o
ce
s
s
v
id
eo
s
ef
f
ec
tiv
ely
.
T
h
er
e
f
o
r
e,
th
e
R
asp
b
er
r
y
Pi
3
is
ty
p
ically
b
etter
s
u
ited
f
o
r
o
u
r
s
y
s
tem
an
d
p
r
o
v
id
es
b
etter
p
er
f
o
r
m
an
ce
f
o
r
v
i
d
eo
ca
p
t
u
r
i
n
g
an
d
s
tr
ea
m
in
g
.
3
.
1
.
2
.
P
i
web
ca
m
er
a
T
h
e
Pi
web
ca
m
er
a
is
ch
o
s
e
n
f
o
r
th
e
p
r
o
p
o
s
ed
s
y
s
tem
d
u
e
to
its
co
m
p
atib
ilit
y
an
d
p
u
r
p
o
s
e
-
b
u
ilt
d
esig
n
f
o
r
R
asp
b
er
r
y
Pi.
I
t
is
id
ea
l
f
o
r
tak
in
g
v
id
eo
r
ec
o
r
d
i
n
g
s
.
Af
ter
p
lu
g
g
in
g
th
e
ca
m
er
a
in
to
th
e
Pi
b
o
ar
d
an
d
g
iv
in
g
it
a
f
ew
co
m
m
a
n
d
s
,
u
s
er
s
ca
n
u
tili
ze
th
e
ca
m
er
a
im
m
ed
iately
.
I
t
also
h
as
th
e
ca
p
ab
ilit
y
o
f
ca
p
tu
r
in
g
i
m
ag
es
with
a
r
eso
lu
tio
n
o
f
5
MP.
R
eg
ar
d
in
g
th
e
p
r
ices,
th
is
ca
m
er
a
co
m
es
at
a
v
er
y
l
o
w
co
s
t,
wh
ich
is
ex
ce
llen
t
v
alu
e
co
n
s
id
er
in
g
th
e
f
ea
tu
r
es
an
d
p
er
f
o
r
m
an
ce
it
o
f
f
er
s
.
T
h
e
co
m
m
an
d
th
at
n
ee
d
s
to
ca
p
tu
r
e
v
id
e
o
is
:
−
1280
×
1024
−
−
10
−
−
ℎ
=
100%
1
.
3
.
1
.
3
.
Ra
s
pb
ia
n
R
asp
b
ian
is
a
f
r
ee
o
p
er
atin
g
s
y
s
tem
tailo
r
ed
f
o
r
cu
s
to
m
i
zin
g
o
r
p
r
o
g
r
am
m
in
g
all
R
asp
b
er
r
y
Pi
m
o
d
els.
I
t
co
m
es
p
r
e
-
i
n
s
talled
with
a
v
ar
iety
o
f
ap
p
licatio
n
s
f
o
r
g
en
er
al
u
s
e,
p
r
o
g
r
am
m
i
n
g
,
an
d
e
d
u
ca
tio
n
al
p
u
r
p
o
s
es.
R
asp
b
ian
s
u
p
p
o
r
ts
p
r
o
g
r
am
m
in
g
lan
g
u
ag
es
s
u
c
h
as
Py
th
o
n
,
J
av
a
,
Scr
atch
,
an
d
m
o
r
e.
T
o
u
s
e
R
asp
b
ian
,
o
n
e
wo
u
ld
in
s
tall
it
o
n
to
a
p
e
r
s
o
n
al
c
o
m
p
u
te
r
(
PC
)
o
r
Mic
r
o
SD
ca
r
d
,
a
n
d
th
e
n
in
s
er
t
it
in
t
o
th
e
R
asp
b
er
r
y
Pi.
Af
ter
co
n
n
ec
ti
n
g
a
m
o
n
ito
r
t
o
th
e
Pi
b
o
a
r
d
,
it
f
u
n
ctio
n
s
lik
e
a
n
o
r
m
al
PC
.
R
asp
b
ian
allo
ws
u
s
er
s
to
ea
s
ily
in
s
tall
n
u
m
e
r
o
u
s
s
o
f
twar
e
p
ac
k
ag
es f
r
o
m
its
o
p
en
-
s
o
u
r
ce
s
o
f
twar
e
r
ep
o
s
ito
r
y
at
n
o
co
s
t.
I
n
t
h
is
p
r
o
ject,
Py
th
o
n
is
em
p
l
o
y
ed
t
o
d
ev
elo
p
th
e
n
ec
ess
ar
y
co
d
e
f
o
r
g
ath
er
in
g
d
ata
f
r
o
m
v
ar
i
o
u
s
s
en
s
o
r
s
to
th
e
R
asp
b
er
r
y
Pi an
d
u
p
lo
ad
in
g
it
to
th
e
I
o
T
clo
u
d
.
3
.
1
.
4
.
M
I
RAC
L
MI
R
AC
L
,
wh
ic
h
s
t
an
d
s
f
o
r
m
u
lt
ip
r
e
cisi
o
n
i
n
t
e
g
e
r
a
n
d
r
a
ti
o
n
al
a
r
it
h
m
et
ic
c
r
y
p
t
o
g
r
a
p
h
ic
li
b
r
a
r
y
,
is
a
s
o
f
twa
r
e
l
ib
r
a
r
y
cr
u
cia
l
f
o
r
c
r
y
p
t
o
g
r
ap
h
i
c
o
p
e
r
a
ti
o
n
s
i
n
v
o
lv
in
g
la
r
g
e
n
u
m
b
er
s
.
I
t
e
f
f
ic
ie
n
tl
y
h
a
n
d
les
a
r
i
th
m
e
tic
o
p
e
r
at
io
n
s
o
n
la
r
g
e
in
te
g
e
r
s
an
d
r
at
io
n
als,
m
ak
in
g
i
t
s
u
ita
b
le
f
o
r
i
m
p
le
m
e
n
ti
n
g
a
lg
o
r
it
h
m
s
li
k
e
E
C
C
a
n
d
R
iv
es
t
–
S
h
am
i
r
–
A
d
le
m
a
n
(
R
S
A
)
e
n
c
r
y
p
t
io
n
.
MI
R
AC
L
is
o
p
t
im
i
ze
d
f
o
r
r
es
o
u
r
c
e
-
c
o
n
s
tr
ai
n
e
d
en
v
i
r
o
n
m
en
ts
an
d
s
u
p
p
o
r
ts
v
a
r
i
o
u
s
c
r
y
p
to
g
r
a
p
h
i
c
s
c
h
e
m
es
wh
ile
m
a
in
tai
n
i
n
g
s
ec
u
r
it
y
t
h
r
o
u
g
h
r
eg
u
l
ar
u
p
d
at
es
a
n
d
a
u
d
its
.
I
t
i
s
a
v
it
al
t
o
o
l
f
o
r
i
m
p
le
m
e
n
ti
n
g
h
i
g
h
-
p
r
e
cisi
o
n
cr
y
p
t
o
g
r
ap
h
i
c
al
g
o
r
it
h
m
s
ef
f
i
cie
n
t
ly
a
n
d
s
ec
u
r
el
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
u
r
ve
illa
n
ce
s
ystem
fo
r
en
h
a
n
cin
g
p
u
b
lic
…
(
S
h
r
ee
d
ev
i Ka
r
ep
p
a
S
a
n
a
p
a
n
n
a
v
a
r
)
7203
3
.
1
.
5
.
E
CDSA
a
lg
o
rit
hm
T
h
er
e
ar
e
3
s
tep
s
:
−
Key
g
en
er
atio
n
First,
a
u
s
er
g
en
er
ates
a
k
ey
p
air
co
n
s
is
tin
g
o
f
a
p
r
iv
ate
k
ey
(
)
r
an
d
o
m
ly
s
elec
ted
in
te
g
er
in
th
e
r
an
g
e
[
1
,
n
-
1
]
.
∈
[
1
,
−
1
]
(
1
)
a
n
d
a
c
o
r
r
e
s
p
o
n
d
i
n
g
p
u
b
l
i
c
k
e
y
(
)
w
h
i
c
h
is
c
a
lc
u
l
a
t
e
d
by
m
u
lt
i
p
l
y
i
n
g
t
h
e
b
as
e
p
o
i
n
t
by
t
h
e
p
r
i
v
a
t
e
k
e
y
.
=
(
2
)
Her
e,
th
e
m
u
ltip
licatio
n
r
ef
er
s
to
s
ca
lar
m
u
ltip
licatio
n
o
n
th
e
ellip
tic
cu
r
v
e,
m
ea
n
in
g
tim
es th
e
p
o
in
t
.
−
Sig
n
in
g
T
o
s
ig
n
a
m
ess
ag
e
,
th
e
s
ig
n
er
f
ir
s
t
co
m
p
u
tes
a
u
n
iq
u
e
v
a
lu
e
f
r
o
m
th
e
m
ess
ag
e
u
s
in
g
a
h
ash
f
u
n
ctio
n
.
=
(
)
(
3
)
T
h
en
,
a
r
an
d
o
m
n
u
m
b
er
is
ch
o
s
en
an
d
u
s
ed
to
c
o
m
p
u
te
a
p
o
in
t
(
1
,
1
)
o
n
th
e
ellip
tic
c
u
r
v
e.
T
h
e
-
co
o
r
d
i
n
ate
o
f
(
1
,
1
)
m
o
d
u
lo
th
e
o
r
d
er
o
f
th
e
c
u
r
v
e
g
i
v
es
th
e
co
m
p
o
n
e
n
t
o
f
th
e
s
ig
n
atu
r
e.
T
h
e
co
m
p
o
n
en
t is ca
lcu
lated
as
(
4
)
,
=
−
1
(
+
)
(
4
)
wh
er
e
is
th
e
p
r
iv
ate
k
ey
,
is
th
e
o
r
d
er
o
f
th
e
c
u
r
v
e,
an
d
−
1
is
t
h
e
m
o
d
u
lar
m
u
ltip
licativ
e
in
v
er
s
e
o
f
m
o
d
u
lo
.
−
Ver
if
icatio
n
T
o
v
e
r
if
y
th
e
s
ig
n
atu
r
e,
t
h
e
v
er
if
ier
u
s
es
th
e
s
ig
n
er
'
s
p
u
b
lic
k
ey
,
th
e
m
ess
ag
e
,
an
d
t
h
e
s
ig
n
atu
r
e
(
,
)
.
First,
(
3
)
is
ca
lcu
lated
f
r
o
m
th
e
m
ess
ag
e.
T
h
en
,
=
−
1
(
5
)
1
=
,
2
=
(
6
)
ar
e
ca
lcu
lated
.
Fin
ally
,
th
e
p
o
i
n
t
(
1
,
1
)
=
1
+
2
(
7
)
is
co
m
p
u
ted
,
wh
er
e
is
th
e
b
ase
p
o
in
t
o
f
th
e
cu
r
v
e
.
I
f
(
1
,
1
)
eq
u
als
(
,
1
)
,
th
e
s
ig
n
atu
r
e
is
v
alid
;
o
th
er
wis
e,
it
i
s
in
v
alid
.
3
.
2
.
Sy
s
t
em
a
rc
hite
ct
ure
T
o
ac
q
u
ir
e
p
u
b
li
c
s
p
ac
e
f
o
o
t
ag
e,
a
R
a
s
p
b
er
r
y
Pi
web
c
am
is
u
ti
li
ze
d
,
an
d
th
e
ca
p
tu
r
e
d
v
id
eo
s
ar
e
s
af
eg
u
a
r
d
ed
u
s
in
g
th
e
el
lip
t
ic
cu
r
v
e
d
ig
it
al
s
ig
n
at
u
r
e
a
lg
o
r
i
t
h
m
(
E
C
D
SA
)
to
p
r
ev
en
t
u
n
au
th
o
r
i
ze
d
a
cc
e
s
s
b
y
h
ac
k
e
r
s
o
r
h
ija
ck
e
r
s
.
T
h
e
s
e
s
ec
u
r
e
d
v
id
eo
s
ar
e
th
e
n
tr
an
s
m
it
ted
to
an
au
t
h
o
r
i
ze
d
in
d
i
v
id
u
a
l
wh
o
u
t
i
li
ze
s
th
em
f
o
r
f
o
r
en
s
ic
an
aly
s
is
.
I
n
th
e
ev
en
t o
f
a
n
y
m
al
ic
io
u
s
ac
t
iv
i
ti
e
s
,
th
e
v
id
eo
s
ar
e
s
cr
u
tin
i
ze
d
to
a
s
ce
r
t
ain
th
e
ty
p
e
s
o
f
s
h
a
r
p
o
b
je
ct
s
in
v
o
lv
ed
,
f
ac
il
it
at
ed
b
y
d
e
ep
le
ar
n
i
n
g
m
o
d
e
l
s
s
u
ch
a
s
V
G
G1
6
.
T
h
e
VG
G1
6
m
o
d
e
l
em
p
lo
y
s
co
n
v
o
l
u
t
io
n
a
l
n
eu
r
al
n
etw
o
r
k
s
(
C
N
N
s
)
to
id
en
tif
y
wh
et
h
er
t
h
e
p
er
p
et
r
a
to
r
i
s
w
ie
ld
in
g
a
g
u
n
o
r
a
k
n
if
e
.
I
f
s
u
ch
o
b
je
ct
s
ar
e
d
e
t
ec
t
ed
,
a
n
o
ti
f
i
ca
tio
n
i
s
s
en
t
t
o
th
e
R
a
s
p
b
er
r
y
P
i,
p
r
o
m
p
t
in
g
i
t
to
a
ct
iv
a
te
an
ala
r
m
s
y
s
t
em
.
3
.
3
.
Resea
rc
h in
deep
lea
rn
ing
Ou
r
r
esear
ch
em
p
lo
y
s
a
th
o
r
o
u
g
h
m
et
h
o
d
o
l
o
g
y
,
wh
ich
is
s
y
s
tem
atica
lly
d
iv
id
ed
in
to
t
h
r
ee
d
is
tin
ct
s
ec
tio
n
s
.
T
h
is
s
tr
u
ctu
r
ed
ap
p
r
o
ac
h
allo
ws
f
o
r
a
clea
r
an
d
o
r
g
an
ized
p
r
esen
tatio
n
o
f
o
u
r
r
esear
ch
p
r
o
ce
s
s
.
Fig
u
r
e
4
v
is
u
ally
illu
s
tr
ates
th
ese
s
ec
tio
n
s
,
p
r
o
v
id
in
g
a
d
etai
led
o
v
e
r
v
iew
o
f
h
o
w
ea
ch
co
m
p
o
n
e
n
t
co
n
tr
ib
u
tes
to
o
u
r
o
v
er
all
r
esear
c
h
s
tr
ateg
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
7
1
9
8
-
7
2
1
0
7204
Fig
u
r
e
4
.
T
h
e
f
lo
w
o
f
r
esear
ch
in
d
ee
p
lear
n
in
g
3
.
3
.
1
.
O
bje
ct
det
ec
t
i
o
n
I
n
itially
,
th
e
v
id
eo
s
o
b
tain
ed
f
r
o
m
th
e
R
asp
b
er
r
y
Pi
web
ca
m
er
a
ar
e
f
ed
i
n
to
th
e
s
y
s
tem
.
T
h
ese
v
id
eo
s
ar
e
th
en
d
ec
o
m
p
o
s
ed
in
to
m
u
ltip
le
f
r
am
es,
wh
ich
allo
ws
f
o
r
a
m
o
r
e
g
r
an
u
lar
an
aly
s
is
o
f
th
e
v
is
u
al
d
ata.
On
ce
th
e
v
id
e
o
s
h
av
e
b
e
en
tr
an
s
f
o
r
m
e
d
in
to
f
r
am
es,
a
s
er
ies
o
f
im
ag
e
p
r
ep
r
o
ce
s
s
in
g
s
tep
s
ar
e
ap
p
lied
.
T
h
is
in
clu
d
es
tech
n
iq
u
es
s
u
ch
as
im
ag
e
tr
an
s
latio
n
,
wh
ich
in
v
o
lv
es
s
h
if
tin
g
th
e
p
ix
els
o
f
th
e
im
ag
e
to
a
ce
r
tain
d
ir
ec
tio
n
,
s
h
ea
r
in
g
,
w
h
ich
d
is
to
r
ts
th
e
im
ag
e
alo
n
g
an
a
x
is
,
n
o
r
m
aliza
tio
n
,
wh
ic
h
ad
ju
s
ts
th
e
p
ix
el
v
alu
es
to
a
s
tan
d
ar
d
ized
r
an
g
e
,
an
d
ed
g
e
d
etec
tio
n
,
wh
ich
id
en
tifie
s
th
e
ed
g
es
o
r
b
o
u
n
d
ar
i
es
o
f
o
b
jects
with
in
th
e
im
ag
e
as d
ep
icted
in
Fig
u
r
e
5
.
Fig
u
r
e
5
.
Pre
p
r
o
ce
s
s
ed
im
ag
es
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
u
r
ve
illa
n
ce
s
ystem
fo
r
en
h
a
n
cin
g
p
u
b
lic
…
(
S
h
r
ee
d
ev
i Ka
r
ep
p
a
S
a
n
a
p
a
n
n
a
v
a
r
)
7205
Af
ter
p
r
e
p
r
o
ce
s
s
in
g
,
th
e
s
y
s
te
m
p
r
o
ce
ed
s
to
p
er
f
o
r
m
o
b
ject
d
etec
tio
n
an
d
id
en
tific
atio
n
in
th
e
in
itial
s
ec
tio
n
.
T
h
is
in
v
o
lv
es
d
etec
tin
g
an
d
r
ec
o
g
n
izin
g
o
b
jects
with
in
th
e
f
r
am
es
u
s
in
g
te
ch
n
iq
u
es
s
u
ch
as
co
n
v
o
l
u
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NN
s
)
o
r
o
th
er
m
ac
h
in
e
le
ar
n
in
g
alg
o
r
ith
m
s
.
Ob
ject
d
et
ec
tio
n
id
en
tifie
s
th
e
p
r
esen
ce
an
d
lo
ca
ti
o
n
o
f
o
b
je
cts
in
th
e
f
r
am
es,
w
h
ile
o
b
ject
id
en
tific
atio
n
d
eter
m
i
n
es
wh
a
t
th
o
s
e
o
b
jects
ar
e,
o
f
ten
b
y
m
atch
in
g
th
em
to
p
r
ed
ef
in
e
d
ca
teg
o
r
ies
o
r
class
es.
T
h
is
p
r
o
ce
s
s
en
ab
les
t
h
e
s
y
s
tem
to
e
x
tr
ac
t
m
ea
n
in
g
f
u
l in
f
o
r
m
atio
n
f
r
o
m
t
h
e
v
id
eo
d
ata,
f
ac
ilit
atin
g
task
s
s
u
ch
as su
r
v
eillan
ce
,
tr
ac
k
in
g
,
o
r
a
n
aly
s
is
.
T
ab
le
1
p
r
o
v
id
es
a
c
o
m
p
r
e
h
e
n
s
iv
e
b
r
ea
k
d
o
wn
o
f
t
h
e
v
id
e
o
s
am
p
les
u
tili
ze
d
f
o
r
test
in
g
p
u
r
p
o
s
es.
Sp
ec
if
ically
,
a
to
tal
o
f
s
ev
en
v
id
eo
s
wer
e
d
e
d
icate
d
to
k
n
if
e
d
etec
tio
n
,
co
llectiv
e
ly
s
eg
m
en
ted
in
to
ap
p
r
o
x
im
ately
3
5
,
6
0
0
f
r
am
es,
ea
ch
ex
clu
s
iv
ely
d
ep
ictin
g
k
n
iv
es.
Ad
d
itio
n
ally
,
ei
g
h
t
v
id
eo
s
wer
e
allo
ca
ted
f
o
r
g
u
n
d
etec
tio
n
,
s
eg
m
en
te
d
i
n
to
ap
p
r
o
x
im
ately
4
2
,
3
4
0
f
r
a
m
es,
ea
ch
s
o
lely
co
n
tain
i
n
g
g
u
n
s
.
T
ab
le
1
.
Nu
m
b
er
o
f
an
n
o
tated
f
r
am
es a
n
d
n
u
m
b
e
r
o
f
v
id
eo
s
W
e
a
p
o
n
s
N
o
.
o
f
f
r
a
m
e
s
N
o
.
o
f
v
i
d
e
o
s
K
n
i
f
e
3
5
,
6
0
0
7
G
u
n
4
2
,
3
4
0
8
To
t
a
l
7
7
,
9
4
0
15
3
.
3
.
2
.
Ana
ly
s
is
Af
ter
o
b
ject
id
en
tific
atio
n
,
th
e
f
r
am
es
co
n
tain
in
g
th
e
d
ete
cted
o
b
jects
ar
e
s
u
b
jecte
d
to
a
tr
ain
in
g
p
h
ase.
Du
r
i
n
g
th
is
p
h
ase,
th
e
s
y
s
tem
lear
n
s
f
r
o
m
t
h
e
d
etec
ted
o
b
jects
to
im
p
r
o
v
e
its
ab
ili
ty
to
r
ec
o
g
n
ize
a
n
d
class
if
y
s
im
ilar
o
b
jects
in
th
e
f
u
tu
r
e.
T
h
e
tr
ain
in
g
d
ata,
co
m
p
r
is
in
g
th
e
d
etec
ted
o
b
jects,
is
u
s
ed
to
r
ef
in
e
th
e
s
y
s
tem
'
s
u
n
d
er
s
tan
d
in
g
o
f
o
b
j
ec
t f
ea
tu
r
es a
n
d
ch
ar
ac
ter
is
tics
.
T
h
e
tr
ain
in
g
d
ata
is
th
en
f
ed
in
to
a
class
if
icatio
n
m
o
d
el,
s
u
ch
as
VGG1
6
,
r
ec
o
g
n
ized
f
o
r
its
ef
f
icac
y
in
im
ag
e
class
if
icatio
n
en
d
ea
v
o
r
s
.
VGG1
6
h
as
m
u
ltip
le
lay
er
s
o
f
c
o
n
v
o
lu
tio
n
an
d
p
o
o
lin
g
o
p
er
atio
n
s
t
h
at
lear
n
h
i
er
ar
ch
ical
ch
a
r
ac
ter
is
tics
ex
tr
ac
ted
f
r
o
m
th
e
in
p
u
t
im
ag
es,
en
ab
lin
g
ac
cu
r
ate
cl
ass
if
icatio
n
.
I
n
th
e
tr
ain
in
g
p
r
o
ce
s
s
,
VGG1
6
lear
n
s
to
as
s
o
ciate
th
e
d
etec
ted
o
b
jects with
s
p
ec
if
ic
class
es o
r
ca
teg
o
r
ies.
3
.
3
.
3
.
Act
io
n
Up
o
n
d
etec
tio
n
o
f
o
b
jects
s
u
c
h
as
g
u
n
s
o
r
k
n
iv
es
b
y
t
h
e
m
o
d
el,
a
n
o
tific
atio
n
s
ig
n
al
is
g
e
n
er
ated
a
n
d
tr
an
s
m
itted
to
th
e
R
asp
b
er
r
y
Pi.
Up
o
n
r
ec
eiv
in
g
th
is
n
o
t
if
icatio
n
,
th
e
R
asp
b
er
r
y
Pi
in
itiates
a
r
esp
o
n
s
e
m
ec
h
an
is
m
,
wh
ic
h
in
cl
u
d
es
a
ctiv
atin
g
a
b
u
zz
e
r
to
s
o
u
n
d
a
n
alar
m
.
T
h
is
alar
m
s
er
v
es
a
s
an
aler
t
to
n
o
tif
y
r
elev
an
t p
ar
ties
ab
o
u
t th
e
p
r
esen
ce
o
f
p
o
ten
tially
d
a
n
g
er
o
u
s
o
b
jects.
3
.
4
.
VG
G
1
6
m
o
del
VGG1
6
is
a
co
n
v
o
l
u
tio
n
al
n
eu
r
al
n
etwo
r
k
(
C
NN)
ar
ch
i
tectu
r
e
s
p
ec
if
ically
d
esig
n
ed
f
o
r
im
a
g
e
class
if
icatio
n
an
d
r
ec
o
g
n
itio
n
task
s
.
T
h
is
ar
ch
itectu
r
e
ex
ce
ls
at
id
en
tify
in
g
an
d
ca
teg
o
r
i
zin
g
o
b
jects
with
in
im
ag
es.
Fig
u
r
e
6
d
is
p
lay
s
a
r
an
g
e
o
f
s
am
p
le
im
ag
es
f
r
o
m
t
h
e
s
tan
d
ar
d
c
o
m
m
o
n
o
b
jects
i
n
co
n
tex
t
(
C
OC
O)
d
ataset,
wh
ich
ar
e
f
o
r
m
atted
a
s
".
jp
g
"
f
iles
.
Sp
ec
if
ically
,
Fig
u
r
e
6
(
a)
th
r
o
u
g
h
Fig
u
r
e
6
(
e
)
s
h
o
w
s
am
p
le
im
ag
es
o
f
g
u
n
s
,
wh
ile
Fig
u
r
e
6
(
f
)
t
h
r
o
u
g
h
Fig
u
r
e
6
(
j)
p
r
esen
t sam
p
le
im
ag
es o
f
k
n
iv
es.
(
a)
(
b
)
(
c)
(
d
)
(
e)
(f)
(
g
)
(
h
)
(
i)
(
j)
Fig
u
r
e
6
.
Sam
p
le
im
a
g
es f
r
o
m
g
ath
er
ed
d
ataset
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
7
1
9
8
-
7
2
1
0
7206
T
h
e
ar
ch
itectu
r
e
co
n
s
is
ts
o
f
f
i
v
e
d
is
tin
ct
lay
er
s
,
ea
c
h
s
er
v
i
n
g
a
s
p
ec
if
ic
f
u
n
ctio
n
with
in
th
e
s
y
s
tem
.
T
h
ese
lay
er
s
wo
r
k
to
g
eth
e
r
to
en
s
u
r
e
th
e
s
ea
m
less
o
p
er
atio
n
an
d
in
teg
r
atio
n
o
f
th
e
o
v
er
a
ll
d
esig
n
.
Fig
u
r
e
7
p
r
o
v
id
es a
d
etailed
illu
s
tr
atio
n
o
f
th
ese
lay
er
s
,
h
ig
h
lig
h
tin
g
t
h
eir
in
d
iv
id
u
al
r
o
les an
d
in
ter
a
ctio
n
s
.
−
I
n
p
u
t
lay
e
r
:
VGG1
6
tak
es
an
in
p
u
t
im
a
g
e
o
f
f
ix
ed
s
i
ze
(
u
s
u
ally
2
2
4
×
2
2
4
p
ix
els)
,
b
u
t
we
tak
e
300
×
3
0
0
p
i
x
els.
−
C
o
n
v
o
lu
tio
n
lay
er
s
:
VGG1
6
co
m
p
r
is
es
o
f
1
3
c
o
n
v
o
lu
tio
n
al
lay
er
s
,
ea
ch
s
u
cc
ee
d
ed
b
y
a
R
eL
U
ac
tiv
atio
n
f
u
n
ctio
n
an
d
3
×
3
co
n
v
o
lu
tio
n
al
f
ilter
s
.
T
h
ese
co
n
v
o
lu
tio
n
al
lay
er
s
ar
e
s
tr
u
ctu
r
ed
to
d
etec
t v
ar
io
u
s
f
ea
tu
r
es
in
th
e
in
p
u
t
im
ag
e,
s
u
ch
as
ed
g
es,
tex
tu
r
es,
an
d
s
h
ap
es,
at
d
if
f
er
en
t
lev
els
o
f
ab
s
tr
ac
tio
n
.
R
eL
U
is
m
ath
em
atica
lly
d
ef
in
e
d
as
,
(
)
=
ma
x
(
0
,
)
w
h
er
e
is
th
e
in
p
u
t to
th
e
f
u
n
ct
io
n
.
−
Po
o
lin
g
l
ay
er
s
:
Af
ter
s
o
m
e
c
o
n
v
o
lu
ti
o
n
al
la
y
er
s
,
VGG1
6
em
p
lo
y
s
m
a
x
-
p
o
o
lin
g
lay
er
s
with
2
×
2
f
ilter
s
an
d
a
s
tr
id
e
o
f
2
.
Po
o
lin
g
lay
er
s
r
ed
u
ce
th
e
s
p
atial
d
im
e
n
s
io
n
s
o
f
th
e
f
ea
tu
r
e
m
a
p
s
,
m
a
k
in
g
th
e
n
etwo
r
k
m
o
r
e
co
m
p
u
tatio
n
ally
ef
f
icien
t a
n
d
r
ed
u
cin
g
t
h
e
r
is
k
o
f
o
v
e
r
f
itti
n
g
.
−
Fu
lly
co
n
n
ec
ted
lay
er
s
:
VGG1
6
co
n
tain
s
th
r
ee
f
u
lly
c
o
n
n
ec
ted
lay
er
s
f
o
llo
wed
b
y
R
eL
U
ac
tiv
atio
n
f
u
n
ctio
n
s
.
T
h
e
f
ir
s
t
two
f
u
lly
co
n
n
e
cted
lay
e
r
s
h
av
e
4
,
0
9
6
n
eu
r
o
n
s
ea
ch
,
an
d
th
e
th
ir
d
f
u
lly
co
n
n
ec
ted
lay
er
h
as
1
,
0
0
0
n
e
u
r
o
n
s
,
wh
i
ch
co
r
r
esp
o
n
d
to
th
e
1
,
0
0
0
c
lass
es
in
th
e
d
ataset
VGG1
6
was
o
r
ig
in
ally
tr
ain
ed
o
n
.
−
Ou
tp
u
t
lay
er
:
T
h
e
o
u
tp
u
t
lay
e
r
o
f
VGG1
6
u
s
es
th
e
s
o
f
tm
ax
ac
tiv
atio
n
f
u
n
ctio
n
,
wh
ich
co
n
v
er
ts
th
e
f
in
al
lay
er
'
s
o
u
tp
u
ts
in
to
p
r
o
b
ab
ilit
ies.
T
h
is
co
n
v
o
lu
tio
n
al
lay
er
is
r
esp
o
n
s
ib
le
f
o
r
th
e
d
etec
tio
n
o
f
wea
p
o
n
s
with
in
th
e
in
p
u
t
im
a
g
es.
T
h
e
r
esu
ltan
t
o
u
tp
u
t
f
ea
tu
r
e
m
ap
s
,
wh
ich
h
ig
h
lig
h
t
t
h
e
p
r
esen
ce
o
f
d
etec
ted
wea
p
o
n
f
ea
tu
r
es,
ar
e
d
e
p
icted
in
Fig
u
r
es 8
(
a)
a
n
d
8
(
b
)
.
Fig
u
r
e
7
.
VGG
1
6
co
n
v
o
lu
tio
n
n
eu
r
al
n
etwo
r
k
Fig
u
r
e
8
.
Ou
t
p
u
t o
f
m
alicio
u
s
o
b
ject
d
etec
ted
(
a
)
k
n
i
f
e
an
d
(
b
)
g
u
n
3
.
5
.
I
nte
rf
a
cing
s
y
s
t
em
s
:
t
ra
ns
m
i
t
t
ing
pa
ck
et
s
t
o
Ra
s
pb
er
ry
P
i
W
ir
esh
ar
k
ca
n
in
d
ee
d
b
e
u
s
ed
to
an
aly
ze
th
e
n
etwo
r
k
tr
a
f
f
i
c
b
etwe
en
a
s
y
s
tem
an
d
a
R
asp
b
er
r
y
Pi.
B
y
ca
p
tu
r
in
g
a
n
d
in
s
p
ec
tin
g
th
e
p
ac
k
ets s
en
t b
etwe
en
th
ese
d
ev
ices,
y
o
u
ca
n
tr
o
u
b
lesh
o
o
t c
o
n
n
ec
tiv
ity
is
s
u
es,
an
aly
ze
p
r
o
t
o
co
ls
,
an
d
e
n
s
u
r
e
th
e
s
ec
u
r
ity
o
f
th
e
c
o
m
m
u
n
icatio
n
.
T
h
is
ca
n
b
e
p
ar
ticu
lar
ly
u
s
ef
u
l
wh
en
s
ettin
g
u
p
n
etwo
r
k
ed
a
p
p
licatio
n
s
o
r
d
iag
n
o
s
in
g
n
etwo
r
k
-
r
elate
d
p
r
o
b
le
m
s
in
p
r
o
jects
in
v
o
lv
in
g
R
asp
b
er
r
y
Pi
.
R
ec
eiv
in
g
th
e
s
im
p
le
s
er
v
ice
d
is
co
v
er
y
p
r
o
to
c
o
l
(
SS
DP)
p
a
ck
ets
f
r
o
m
t
h
e
s
y
s
tem
I
P
ad
d
r
ess
to
R
asp
b
er
r
y
Pi,
as
illu
s
tr
ated
in
Fig
u
r
e
9
.
W
h
en
th
e
s
y
s
tem
d
etec
ts
g
u
n
s
o
r
k
n
iv
es,
it
tr
ig
g
er
s
a
n
o
tific
atio
n
to
b
e
s
en
t
to
t
h
e
R
asp
b
er
r
y
Pi,
wh
ich
in
tu
r
n
ac
tiv
ates a
n
aler
t b
u
zz
to
t
h
e
p
u
b
lic.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
d
ee
p
lea
r
n
in
g
-
b
a
s
ed
s
u
r
ve
illa
n
ce
s
ystem
fo
r
en
h
a
n
cin
g
p
u
b
lic
…
(
S
h
r
ee
d
ev
i Ka
r
ep
p
a
S
a
n
a
p
a
n
n
a
v
a
r
)
7207
Fig
u
r
e
9
.
Pack
ets f
r
o
m
th
e
s
y
s
tem
to
R
asp
b
er
r
y
Pi
4.
RE
SU
L
T
S
T
h
e
o
u
tp
u
t
f
r
o
m
th
e
VGG1
6
m
o
d
el
is
au
to
m
atica
lly
s
av
ed
as
a
.
tx
t
f
ile
o
n
t
h
e
lo
ca
l
d
ev
ic
e
as
s
h
o
wn
in
Fig
u
r
e
9
.
Su
b
s
eq
u
e
n
tly
,
th
i
s
tex
t
f
ile
is
tr
an
s
m
itted
f
r
o
m
th
e
clien
t
s
id
e,
r
u
n
n
in
g
Py
th
o
n
3
.
1
2
.
3
o
n
a
lo
ca
l
d
ev
ice,
t
o
th
e
s
er
v
er
,
wh
ich
is
h
o
s
ted
o
n
a
R
asp
b
er
r
y
Pi.
I
f
t
h
e
s
er
v
er
r
ec
eiv
es d
ata
in
d
icatin
g
a
n
in
te
g
er
v
alu
e
o
f
1
,
s
ig
n
if
y
in
g
th
e
d
etec
tio
n
o
f
a
Gu
n
o
r
Kn
if
e,
it
tr
ig
g
er
s
an
aler
t
v
ia
a
b
u
zz
e
r
co
n
n
ec
ted
to
th
e
R
asp
b
er
r
y
Pi.
Oth
er
wis
e,
it d
is
p
lay
s
th
e
m
ess
ag
e
“
YOU
AR
E
I
N
SA
F
E
.
”
T
h
e
r
esu
lts
ar
e
s
h
o
wn
b
elo
w.
W
h
en
f
ir
ea
r
m
s
ar
e
d
etec
ted
,
th
e
r
esu
lts
will
b
e
d
is
p
lay
ed
o
n
th
e
clien
t
-
s
id
e
in
ter
f
ac
e,
p
r
o
v
id
in
g
im
m
ed
iate
f
ee
d
b
ac
k
to
u
s
er
s
.
Simu
ltan
eo
u
s
ly
,
th
is
in
f
o
r
m
atio
n
will
also
b
e
s
h
o
wn
o
n
th
e
s
er
v
er
-
s
id
e
in
ter
f
ac
e,
e
n
s
u
r
in
g
th
at
th
e
d
a
ta
is
av
ailab
le
f
o
r
b
ac
k
en
d
p
r
o
ce
s
s
in
g
an
d
m
o
n
ito
r
in
g
.
Fig
u
r
e
1
0
d
em
o
n
s
tr
ates
h
o
w
th
e
r
esu
lts
ap
p
ea
r
o
n
th
e
clien
t
-
s
id
e
in
ter
f
ac
e,
w
h
ile
Fig
u
r
e
1
1
illu
s
tr
ates
th
eir
p
r
esen
tatio
n
o
n
th
e
s
er
v
er
-
s
id
e
in
ter
f
ac
e.
I
f
n
o
f
ir
ea
r
m
s
a
r
e
d
etec
ted
,
a
n
o
tific
atio
n
m
ess
ag
e
will
b
e
d
is
p
lay
ed
o
n
th
e
clien
t
-
s
id
e
i
n
ter
f
ac
e
t
o
in
f
o
r
m
u
s
er
s
o
f
th
e
a
b
s
en
ce
o
f
th
r
ea
ts
.
At
th
e
s
am
e
tim
e,
th
is
s
tatu
s
m
es
s
ag
e
will a
l
s
o
b
e
s
h
o
wn
o
n
th
e
s
er
v
er
-
s
id
e
in
ter
f
ac
e,
en
s
u
r
in
g
th
at
b
ac
k
en
d
s
y
s
tem
s
ar
e
u
p
d
ated
a
cc
o
r
d
in
g
l
y
.
Fig
u
r
e
1
2
illu
s
tr
at
es
h
o
w
th
is
m
es
s
ag
e
ap
p
ea
r
s
o
n
t
h
e
clien
t
-
s
id
e
in
ter
f
ac
e,
wh
ile
Fig
u
r
e
1
3
s
h
o
ws its
p
r
esen
tatio
n
o
n
t
h
e
s
er
v
er
-
s
i
d
e
in
ter
f
ac
e.
Fig
u
r
e
1
0
.
Dete
cted
f
ir
ea
r
m
s
o
n
clien
t
-
s
id
e
o
u
tp
u
t
Fig
u
r
e
1
1
.
Dete
cted
f
ir
ea
r
m
s
o
n
s
er
v
er
-
s
id
e
o
u
tp
u
t
Evaluation Warning : The document was created with Spire.PDF for Python.