Emergence and development of digital watermarking technology has close relation with demands of daily life. Multiscale geometric analysis (MGA) provides a new tool and approach for study of digital watermarking. In compare to wavelet transform, MGA is more effective methods of image sparse representation. Using MGA approach, digital watermarking algorithm usually has better performance.A novel digital watermarking algorithm based on nonsubsampled contourlet transform (NSCT) and a novel digital watermarking algorithm in Curvelet domain is presented in this paper. Scrambling and encrypting the watermarking information using chaos is able to enhance the security of digital watermarking, while the anti-attack ability of shear and other attacks is improved. To make better use of image masking, fuzzy control method is used to control watermarking embedding strength and it can improve the performance of watermarking algorithm.The experimental results show that the algorithm proposed in this paper has good security, robustness and invisibility, and it can effectively resist the JPEG compression, filtering, cropping and other attacks. |