Cryptography Reference
In-Depth Information
Table 11.1.
Experimental Results for the Pepper and F16 Images.
Operations
Image
Pepper
F16
Embedding
29.879
29.901
(PSNR in dB)
Robust
Semi
Robust
Semi
(NHS)
-Fragile
(NHS)
-Fragile
(NHS)
(NHS)
Extraction (no attack)
1.000
1.000
1.000
1.000
JPEG (QF = 100%)
0.998
0.955
0.990
0.940
JPEG (QF = 80%)
0.969
0.861
0.964
0.824
JPEG (QF = 50%)
0.951
0.705
0.949
0.683
JPEG (QF = 30%)
0.939
0.388
0.937
0.387
VQ (Codebook 1)
1.000
1.000
1.000
1.000
VQ (Codebook 2)
0.901
0.518
0.889
0.500
VQ (Codebook 3)
0.830
0.049
0.791
0.089
VQ (Codebook 4)
0.912
0.177
0.890
0.189
Image cropping
0.789
0.948
0.857
0.921
Median filtering
0.871
0.132
0.888
0.124
Blurring
0.925
0.634
0.932
0.615
High-Pass filtering
0.614
0.049
0.627
0.093
Contrast Enhancement
0.918
0.254
0.920
0.215
Brightness Enhancement
0.701
0.045
0.713
0.035
Rotation by 0.5
◦
clockwise
0.801
0.069
0.732
0.143
Rotation by 0.5
◦
counterclockwise 0.782
0.058
0.693
0.120
Rotation by 3
◦
clockwise
0.621
0.091
0.604
0.098
Rotation by 3
◦
counterclockwise
0.601
0.089
0.598
0.084
ond stage using a simple index constrained method. Although the encoded
indices of the attacked watermarked image may be very different from the
original ones, the variance of neighboring indices does not vary a great deal.
The first-stage watermarking method is therefore robust. The second-stage
watermarking method is based on an index constrained codeword search pro-
cedure, in which the index is modified according to the bit to be embedded.
Any change in the encoded indices may introduce the change in the extracted
watermark bit. In other words, the second-stage watermarking method can
tolerate few modifications, which means is fragile to most intentional attacks.
Experimental results demonstrate that the proposed method can be used for
copyright protection by extracting the first-stage watermark. It can also be
used for image authentication by extracting the second-stage watermark.
Compared with existing multipurpose watermarking algorithms, the ad-
vantages of the proposed algorithm are as follows:
1) The proposed algorithm can tolerate rotation attacks with having larger
angles.
