Steganography in Halftone Images - Visual Cryptography and Secret Image Sharing - page 355

FIGURE 13.9
DHCED-generated Y 2 (T = 10) of Lena with respect to X in Figure 13.2, W
in Figure 13.5, and Y 1 in Figure 13.4.
Comparing DHSED and DHCED, it can be observed that DHSED in Fig-
ure 13.6 has strong boundary artifacts along the embedded watermark es-
pecially on the right and bottom edges of W. With DHCED, the boundary
artifacts are reduced considerably. In addition, the watermark in Figure 13.10
is significantly more visible than Figure 13.7 in terms of contrast, when Y 1
and Y 2 are overlaid.
DHCED for the case of different X 1 and X 2 is also simulated. Here Lena
is used as X 1 and Pepper is used as X 2 . The Y 1 with respect to X 1 is simply
the one in Figure 13.4. The DHCED generated Y 2 from X 2 with respect to Y 1
and W is shown in Figure 13.12. A threshold of T = 10 is used. The overlaid
image of Y 1 and Y 2 is shown in Figure 13.13. As expected, traces of both
Lena and Pepper can be observed in the overlaid image. More importantly,
the watermark W is revealed also, though the contrast of W in Figure 13.13
is not as good as in Figure 13.10.
13.5 Performance Analysis
In this section, we give an indepth analysis of both DHSED and DHCED. We
will show theoretically why DHCED has better performance than DHSED.