The Application of an Improved Baker Map in Image Scrambling and Watermarking

Thanks to the exceptionally good properties, chaotic systems have been widely studied and developed in image scrambling and watermarking algorithms in recent years. In this paper, the chaotic standard Baker map is improved and employed to scramble the pixel positions of host image and watermark image in the preprocessing of watermarking. Experimental results show that the scrambling algorithm based on the improved Baker map can greatly disorder the pixel positions to enhance the robustness of the watermarking scheme. The watermarking is performed in the spatial domain, in which one sign pattern matrix is used to determine those pixels in the host image to embed the watermark bits. Such a performance spreads the watermark bits in the edge of the host image, and therefore increases the transparency of the watermark.