Cropping, rotation and scaling invariant LBX interleaved voice-in-image watermarking

  • Authors:

  • Sung Shik Koh;Chung Hwa Kim

  • Affiliations:

  • Dept. of Information and Communication Engineering, Osaka City University, Osaka, Japan;Dept. of Electronics, Chosun University, Gwangju, Korea

  • Venue:
  • AIS'04 Proceedings of the 13th international conference on AI, Simulation, and Planning in High Autonomy Systems
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

The rapid development of digital media and communication network urgently has highlighted the need of data certification technology to protect IPR (Intellectual property rights). This paper proposed a new watermarking method for embedding the owner's voice signal using our LBX (Linear Bit-eXpansion) interleaving. This method uses a voice signal as a watermark to be embedded, which makes it quite useful in claiming ownership, and has the advantage of restoring a voice signal that has been modified and removed by image removing attacks by applying our LBX interleaving. Three basic stages of this watermarking include: 1) Encode the analogue owner's voice signal by PCM and create new digital voice watermark; 2) Interleave a voice watermark by LBX; and 3) Embed the interleaved voice watermark in the low frequency band on DHWT (Discrete Haar Wavelet Transform) of the blue and red channels of the color image. Therefore the resulting model can be used to maximize the robustness against attacks for removing a part of image such as cropping, rotation and scaling, because de-interleaver can correct the modified watermark information.