A generalised model for distortion performance analysis of wavelet based watermarking

Deepayan Bhowmik*, Charith Abhayaratne

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingChapter (peer-reviewed)

7 Citations (Scopus)


A model for embedding distortion performance for wavelet, based watermarking is presented in this paper. Firstly wavelet based watermarking schemes are generalised into a, single common framework. Then a mathematical approach has been made to find the relationship between distortion performance metrics and the watermark embedding parameters. The derived model shows that for wavelet based watermarking schemes the sum of energy of the selected wavelet coefficients to be modified is directly proportional to the distortion performance (the mean square error) measured ill the pixel domain. The propositions are made using the energy conservation theorem between input, signal and transform domain coefficients for orthonormal wavelet bases. Such all analysis is useful to choose the wavelet coefficients during watermark embedding procedure and to find suitable input parameters such as wavelet kernel or the choice of subband.

Original languageEnglish
Title of host publicationDigital Watermarking
Subtitle of host publication7th International Workshop, IWDW 2008, Busan, Korea, November 10-12, 2008. Selected Papers
EditorsHyoung-Joong Kim, Stefan Katzenbeisser, Anthony T S Ho
Number of pages16
ISBN (Electronic)978-3-642-04438-0
ISBN (Print)978-3-642-04437-3
Publication statusPublished - 2009
Event7th International Workshop on Digital Watermarking - Busan, Korea, Democratic People's Republic of
Duration: 10 Nov 200812 Nov 2008

Publication series

NameLecture Notes in Computer Science
ISSN (Print)0302-9743


Conference7th International Workshop on Digital Watermarking
Abbreviated titleIWDW 2008
Country/TerritoryKorea, Democratic People's Republic of


  • Watermarking
  • wavelet transforms
  • distortion performance


Dive into the research topics of 'A generalised model for distortion performance analysis of wavelet based watermarking'. Together they form a unique fingerprint.

Cite this