Multiple-description framework for oblivious watermarking
R. Chandramouli, Benjamin Myer Graubard, Colin R. Richmond · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2001
We propose for the first time a multiple description framework for oblivious watermarking. Parallels between multiple description source coding and the watermarking are drawn. An information theoretic definition of the problem is given. A spread-spectrum watermarking algorithm for DCT based multiple descriptions is described. Performance of the proposed framework for various attack channels such as additive white Gaussian noise, MPEG compression, and random bit error channels shows that the proposed method performs reasonably well compared to non-oblivious schemes.