Information Technology Reference
In-Depth Information
where
σ
P
is the magnitude of peak signal and
S
is the size of host data. In the experi-
ment,
σ
P
=255 and
S
=256×256.
With the same amount of data embedded, robustness and imperceptibility were ef-
fectively increased by using the proposed approach compared to the original method
as shown in Fig.3 where the embedded data are 16 alphanumeric characters.
Alternatively, with the same robustness and imperceptibility, a larger embedding
capacity can be achieved by using the proposed method. If robustness and impercep-
tibility of the original scheme are kept at the same level as shown by the solid line in
Figure 3, both the numbers of bits and characters embedded using the old method can
be calculated. The results are listed in Table 2, from which one can see that the new
method provides a higher embedding capacity.
25
20
15
10
5
0
44.5
45
45.5
46
46.5
Fig. 3.
Comparison of robustness and imperceptibility using original (dashed line) and pro-
posed (solid line) approaches with 16 alphanumeric characters embedded
Peak-signal-to-watermark power ratio (dB)
Table 2.
Embedding capacities with different imperceptibility and robustness
0.045
CER (%)
2.50
1.11
0.46
0.16
0.011
0.0037
PSWR (dB)
46.0
45.5
45.0
44.5
44.0
43.5
43.0
Proposed
method
(
K
M
)
16
16
16
16
16
16
16
Number of
embedded
characters
Original
method
(
K
S
)
8
7
7
6
6
5
5
2.00
2.29
2.29
2.67
2.67
3.20
3.20
K
=
K
M
/
K
S
5.3
Simulation Results
A low rate of character extraction error is required in practical applications. In the
experiment,
α
Μ
= 2.72 was used to give a CER = 1.0% under AGWN interference at
