Information Technology Reference

In-Depth Information

Here,
CTPT
ca
(
char
i
(
LP
)
,char
i
(
LP
)
,w,t
) stands for the CTPT value of

camera
ca
under the condition of
char
i
(
LP
) which is the original character,

char
i
(
LP
)whichis
char
i
(
LP
) transformed into, weather
w
and time period
t
.

Based on this assumption, we propose a correction algorithm, namely
SLPs

Correction, and there are two main steps including get potential
CLPs
and

correct
SLPs
. Basic steps are illustrated in Algorithm 2.

Algorithm 2.
SLPs
Correction

Input:
SLPs
of each camera
ca ∈ CA
, the limit of
k
when searching
k
-degree neigh-

bors, the time
t
at which the
SPL
being captured, the continuous occurrence limit

ʸ
,andtheinterval
ʔt
between the camera in question and the
k
-degree neighbors

Output:
Output
CLPs
for each camera
ca ∈ CA
and its corresponding
SLPs
in
ca

1:
for
each
ca ∈ CA
do

2:

// get potential CLPs

3:

potentialCLPs
=
∅
;

for
i ₐ
1to
k
do

4:

5:

for
each
LP
recognized by
ca
do

6:

if
ʸ

≥ ʸ −
1in
ca
's 1,2,...,
k
-degree neighbors in interval
ʔt
then

7:

Add

LP

to

potentialCLPs

;

8:

end if

9:

end for

10:

end for

11:

//correct SLPs

12:

for
each
slp ∈ SLPs
do

13:

if
potentialCLPs
equals
∅
then

14:

Mark
slp
as
Exp
;

15:

else

Select most approximate
slp
by maximizing
sim
(
slp,slp
,w,t
)
,slp
∈

potentialCLPs
;

16:

as
CLP
;

17:

Mark
slp

18:
end if

19:
end for

20: Output
CLPs
and its corresponding
SLPs
;

21:
end for

CLPs

Accuracy Promotion.
To get a higher accuracy of
CLPs
, data fusion

of RFID and ALPR is applied in the process of
CLPs
Accuracy Promotion.

Since RFID records are more accurate than ALPR to be convincible, those
CLPs

whose original data both have ALPR
LP
records and RFID
LP
records could

transform into the results recognized by RFID directly. In a way, promotion of

CLPs
accuracy would be acquired to enhance the correction result. Algorithm

3,
CLPs
Accuracy Promotion, shows the detail of how data fusion improves the

CLPs
accuracy.

4 Experiment and Evaluation on In-Memory Database

In this section, experiments are designed to assess the veracity and time con-

suming of HelpMe. To evaluate the performance of HelpMe in veracity, ALPR