Geometrically invariant image watermarking in the DWT domain

Shijun Xiang, Hyoung Joong Kim

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)


Watermark resistance to both geometric attacks and lossy compressions is a fundamental issue in the image watermarking community. In this paper, we propose a DWT (Discrete Wavelet Transform) based watermarking scheme for such a challenging problem. Watermark resistance to geometric deformations is achieved by using the invariance of the histogram shape. In both theoretical analysis and experimental way, we show that the invariance can be extended to the DWT domain thanks to the time-frequency localization property of DWT. Consequently, we achieve the goal to embed a geometrically invariant watermark into the low-frequency sub-band of DWT in such a way that the watermark is not only invariant to various geometric transforms, but also robust to common image processing operations. Extensive simulation results demonstrate the superiority of the proposed watermark strategy due to the use of the histogram shape invariance combined with the DWT technique.

Original languageEnglish
Title of host publicationInformation Security Applications - 8th International Workshop, WISA 2007, Revised Selected Papers
PublisherSpringer Verlag
Number of pages15
ISBN (Print)354077534X, 9783540775348
Publication statusPublished - 2007
Event8th International Workshop on Information Security Applications, WISA 2007 - Jeju Island, Korea, Republic of
Duration: 2007 Aug 272007 Aug 29

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4867 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other8th International Workshop on Information Security Applications, WISA 2007
Country/TerritoryKorea, Republic of
CityJeju Island

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Geometrically invariant image watermarking in the DWT domain'. Together they form a unique fingerprint.

Cite this