Salt-and-pepper image watermarking system for IHC evaluation criteria

This paper presents an image watermarking system for the IHC evaluation criteria (ver.3) which require tolerance against both compression and cropping. To meet the requirement, our system repeatedly embeds watermark information into the DCT domain with significantly high watermark strength. On the other hand, to keep high image quality, only one DCT coefficient is used for each watermark bit. Thus, the system adds spike noise (or salt-and-pepper noise) to the DCT domain at embedding. The experimental results show that the watermarking system satisfies the IHC criteria ver.3 and achieves high performance in the contest categories called highest tolerance and highest image quality. Specifically, in the highest tolerance category, the embedded watermark information is detected without error even against compression reducing the file size by 0.5 of its original size. On the other hand, in the highest image quality, PSNR is not lower than 41 dB without error against the specified compression ratio, namely, 4%.

[1]  Mauro Barni,et al.  A new decoder for the optimum recovery of nonadditive watermarks , 2001, IEEE Trans. Image Process..

[2]  Panagiotis Tsakalides,et al.  Hidden messages in heavy-tails: DCT-domain watermark detection using alpha-stable models , 2005, IEEE Transactions on Multimedia.

[3]  Michael G. Strintzis,et al.  Locally optimum nonlinearities for DCT watermark detection , 2004, IEEE Transactions on Image Processing.

[4]  Juan R. Hern,et al.  Statistical Analysis of Watermarking Schemes for Copyright Protection of Images , 1999 .

[5]  Ioannis Pitas,et al.  Circularly symmetric watermark embedding in 2-D DFT domain , 2001, IEEE Trans. Image Process..

[6]  Josep Domingo-Ferrer,et al.  Spatial-Domain Image Watermarking Robust against Compression, Filtering, Cropping, and Scaling , 2000, ISW.

[7]  Fernando Pérez-González,et al.  DCT-domain watermarking techniques for still images: detector performance analysis and a new structure , 2000, IEEE Trans. Image Process..

[8]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[9]  R. HernandezJ.,et al.  DCT-domain watermarking techniques for still images , 2000 .

[10]  Fernando Pérez-González,et al.  Approaching the capacity limit in image watermarking: a perspective on coding techniques for data hiding applications , 2001, Signal Process..

[11]  Ioannis Pitas,et al.  Asymptotically optimal detection for additive watermarking in the DCT and DWT domains , 2003, IEEE Trans. Image Process..