A reversible image authentication scheme based on chaotic fragile watermark

Di Xiao, Frank Y. Shih

Research output: Contribution to journalArticlepeer-review

6 Scopus citations

Abstract

Although a variety of watermark-based image authentication schemes have been proposed so far, most of them have to leave some permanent distortions in the original image for embedding the watermark information. In this paper, we propose a novel reversible image authentication scheme to allow complete recovery of the cover image after the extraction of the embedded watermark. In the embedding process, the original image is partitioned into blocks; the chaotic watermark of each block is computed and embedded into the block itself in a reversible way. In the extracting process, the watermarked image is partitioned into blocks; the embedded watermark of each block is extracted and the block itself is fully recovered. Then we compute the new chaotic watermark of each recovered block and perform comparison with the extracted one. Therefore, we can detect whether the original image is modified. If it is, then the modification can be located. Besides, by the use of chaotic watermark, the complicated nonlinear and sensitive dependence within image feature, secret key, and watermark is established to ensure the watermarking security. Theoretical analysis and computer simulation have verified the successful achievement of the proposed image authentication scheme.

Original languageEnglish (US)
Pages (from-to)4731-4742
Number of pages12
JournalInternational Journal of Innovative Computing, Information and Control
Volume6
Issue number10
StatePublished - Oct 2010

All Science Journal Classification (ASJC) codes

  • Software
  • Theoretical Computer Science
  • Information Systems
  • Computational Theory and Mathematics

Keywords

  • Authentication
  • Chaos
  • Fragile watermark
  • Hash
  • Reversible watermark

Fingerprint

Dive into the research topics of 'A reversible image authentication scheme based on chaotic fragile watermark'. Together they form a unique fingerprint.

Cite this