Towards the public but noninvertible watermarking

  • Authors:
  • Xianfeng Zhao;Yingxia Dai;Dengguo Feng

  • Affiliations:
  • The State Key Laboratory of Information Security, Graduate School of Chinese Academy of Sciences, Beijing, P. R. China;The State Key Laboratory of Information Security, Graduate School of Chinese Academy of Sciences, Beijing, P. R. China;The State Key Laboratory of Information Security, Graduate School of Chinese Academy of Sciences, Beijing, P. R. China

  • Venue:
  • IWDW'04 Proceedings of the Third international conference on Digital Watermarking
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

The public watermarking schemes, which do not need the originals in ownership verification, have been proved invertible in nature. However, this paper is to demonstrate that some public but noninvertible schemes exist in the cases of the ill-posed linear embedding or nonlinear embedding. Since attackers have to solve an ill-posed linear or nonlinear system, they have great difficulty dividing a released version into their claimed original data and scaled watermarks, and in the meantime making the latter be the particular adaptive results based on the former. Since the inverted solutions are drastically perturbed and perceptually unacceptable, they can fail to cheat the private schemes. Furthermore, because the owner has the embedded watermark which jointly defines the embedding system, he or she can do the reverse engineering more precisely and ultimately differentiate himself or herself from the attackers in a public scheme, achieving the public but noninvertible schemes.