Cryptography Reference
In-Depth Information
0.25
0.25
0.225
0.225
0.2
0.2
0.175
0.175
0.15
0.15
0.125
0.125
Estimated BER
Measured BER
0.1
0.1
Estimated BER
Measured BER
0.075
0.075
0.05
0.05
0.025
0.025
Rate of
improvement
0
0
Rate of
improvement
S
1112131415161718191101111
Number of accumulations
S
1112131415161718191101111
Number of accumulations
opt
opt
(a) Effective case
(3 Mbps, detecting point 9)
(b) Non-effective case
(3 Mbps, detecting point 14)
Fig. 7.16. Transitions of BERs.
Example transitions of the BERs estimated using the statistically adaptive
technique (corresponding to Fig. 7.11) are shown in Figs. 7.16(a) and 7.16(b),
where the horizontal axes represent the number of accumulations (the order
of the accumulation follows the ascending order of the estimated BERs), and
the vertical axes represent the estimated and measured BERs at the detect-
ing points. From these transitions we can see that the estimated BERs were
roughly consistent with the experimentally measured ones. In Fig. 7.16(a)
(3 Mbps, detecting point 9), the BER tended to increase with the number
of accumulations, and the rate of improvement was about 2.5%. For all bit
rates, similar trends were found for detecting points 7 through 13, where most
parts of the neighboring frames were messy or moving. We can infer that the
strengths of WMs in regions having various picture properties are affected by
MPEG-2 encoding so that the BERs of the regions increased with the num-
ber of accumulations as a result of the sorting operation of the statistically
adaptive detection technique. In Fig. 7.16(b) (3 Mbps, detecting point 14), on
the other hand, the plot of BER against the number of accumulations is flat,
and the rate of improvement is nearly 0. For all bit rates, similar trends were
found for detecting points 1 through 6, 14, and 15, where most parts of the
neighboring frames were static. We can infer that the strengths of WMs in
the regions were not affected by MPEG-2 encoding so that the BERs of the
regions were not changed by the sorting operation.
For all the MPEG-2 bit rates evaluated, the statistically adaptive detection
technique produced lower or equal BERs depending on the picture properties
and yielded an average improvement of 9.7%. The statistically adaptive de-
tection technique thus improves watermark detection.
Search WWH ::




Custom Search