Cryptography Reference
In-Depth Information
protection and ownership verification because they are robust to nearly all
kinds of image processing operations. In comparison, fragile or semi-fragile
watermarks [7, 8, 9, 10] are mainly applied to content authentication and
integrity attestation because they are fragile to most modifications. Most ex-
isting watermarking schemes are designed for either copyright protection or
content authentication. However, we should also consider using multiple wa-
termarks in the following cases:
1) Alice sells a digital product or artwork to Bob. Bob wants to know if the
product is authentic or counterfeit. On the other hand Alice wants to know
if Bob or other users have been reselling her product;
2) Alice sends Bob some important commercial information in the form of
multimedia. Bob wants to know where the information is from or who
makes the decision. Bob also wants to know if the information has been
tampered with by others;
3) Alice sends to Bob some secret military information by securely embed-
ding this information into an irrelevant image. Bob wants to know if the
embedded information is authentic.
Thus, the watermarking algorithms that can fulfill multipurpose and are for
this reason alone urgently required.
A visible watermark typically contains a visible message or a company logo
indicating the ownership of the image. A visible watermark can be used to
convey immediate claim of the copyright and prevent the unauthorized com-
mercial use of images. Howsoever robust a visible watermark might be, it can
always be partially or fully removed using software. To provide further protec-
tion for the image work, we give a method to embed an invisible watermark
as a back-up in the host image in addition to the visible watermark.
For the third application, we should use effective retrieval methods to get
the desired images. In general, the visual contents of the images in the data-
base are extracted and are described by multi-dimensional feature vectors.
The feature vectors form a feature database. To retrieve images, users provide
the retrieval system with example images or sketched figures. The similari-
ties/distances between the feature vectors of the query example or sketch and
those of the images in the database are then calculated. The retrieval is also
done with the aid of an indexing scheme. Recent retrieval systems have incor-
porated users relevance feedback to modify the retrieval process. In general,
the copyright protection and image retrieval issues are considered separately.
The last part of this chapter presents a multipurpose watermarking scheme
to solve these two problems simultaneously.
