Reversible Fragile Watermarking for Multichannel Images with High Redundancy Channels
- PDF / 1,529,479 Bytes
- 19 Pages / 439.37 x 666.142 pts Page_size
- 8 Downloads / 193 Views
Reversible Fragile Watermarking for Multichannel Images with High Redundancy Channels Marco Botta 1 & Davide Cavagnino 1 & Victor Pomponiu 1 Received: 13 September 2018 / Revised: 13 March 2020 / Accepted: 22 April 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
The paper presents a methodology to protect the integrity of multichannel images, having some highly redundant channels, by means of a reversible fragile watermarking algorithm. The watermark embedding phase uses a lossless compression method to compress the high redundancy channels, stores the compressed stream into their most significant bits, then embeds a secret fragile watermark by modifying the least significant bits of the high redundancy channels. In case the watermarked image is not modified, the host image can be perfectly reconstructed; otherwise, the modified area can be detected and located with very high probability and the area that has not been forged can be restored as in the original host image. The embedding of the watermark is performed by a Genetic Algorithm in the Karhunen-Loève Transform (KLT) domain: the use of a secret space defined by the KLT guarantees both security of the method and a high sensitivity in the detection of the forged areas. Keywords Reversible watermarking . Fragile watermarking . Multichannel image processing . Image authentication . Genetic algorithm . Karhunen-Loève Transform
1 Introduction Digital data may undergo different kinds of attacks, like malicious modification (i.e. tampering), unauthorized copying and copyright infringement.
* Marco Botta [email protected] Davide Cavagnino [email protected] Victor Pomponiu [email protected]
1
Dipartimento di Informatica, Università degli Studi di Torino, Corso Svizzera 185, 10149 Torino, Italy
Multimedia Tools and Applications 19
In the field of data hiding, digital watermarking (watermarking in the following) is a technology that may be used to protect digital contents, like images, videos and sounds, from the cited security attacks. Basically watermarking embeds a signal W into a digital object I by modifying some parts of it. Depending on the objective of the protection, different watermarking algorithms are developed. Before describing the properties of a watermarking algorithm we introduce a simple model of the whole process. The watermarking process is composed by two distinct phases, sequential and nonoverlapping in time. The first one is the embedding phase, which inserts a watermark signal W into an object I possibly using a secret key K: the output is a watermarked object IW. The 0 second phase is a extraction stage: after receiving I W (a possibly altered watermarked object due to intentional attacks or to various kinds of transmission or storage errors) it returns the watermark signal present W′ and/or a Boolean value stating if the original watermark is present; this phase, in general, needs the secret key K used to embed the watermark and may also require the original image I and watermark W.
Data Loading...
