Cryptography Reference
In-Depth Information
(a)
(b)
The cropped version of the water-
marked image (δ =4)
The
restored
image
block
corre-
sponding to (a)
(c)
(d)
The cropped version of the water-
marked image (δ =8)
The
restored
image
block
corre-
sponding to (c)
Fig. 13.15. The cropped watermarked image and the retrieved image.
In Fig. 13.15, generally δ is equal to 4. If the size of the cropped image
is very small, for example the size being not larger than 3232. To find the
correct cropped position, we can for example select δ = 8 in Fig. 13.15.
In the practical application, the watermarked image can suffer from collage
attack. We use the image ID and the block index as the input to the hash
function. The scheme can e ciently resist collage attack. Fig. 13.16 shows
that our method can e ciently resist collage attack.
This chapter presents a detailed investigation on the development of all
existing lossless data hiding techniques. The mechanism, the merits and the
drawbacks of these techniques are discussed. In this chapter we introduce our
own method for reversible watermarking.
13.5 The Future Research
High-capacity lossless data-embedding algorithms still provide an active re-
search foreground. Robust lossless data-embedding algorithms have attracted
considerable attention from researchers in recent years. Reversible algorithms
such as JPEG compression are capable of retrieving watermarking informa-
tion even if the watermarked images under attack. Shi et al. [18] note that
robust lossless data hiding algorithms may find wide applications in semi-
fragile authentication of JPEG2000 compressed images. Future research work
Search WWH ::




Custom Search