An Algorithm to Improve the Performance of Watermarking Systems

Random permutation and sharpening are basic methods of image processing. In some digital image watermarking schemes, if original image is pre-processed with random permutation before embedding and the watermarked image is pre-processed with sharpening before detecting, the performance of watermarking systems can be improved, that means the images can seem better, or the veracity rate of detection is higher. In this article, this method is validated in Barni’s semi-blind watermarking system. The method is suitable for some semi-blind or blind watermarking systems, whether in spatial domain or in other transform domain.