Robust image watermarking in contourlet domain using multi objective genetic algorithm

This paper presents a novel digital watermarking algorithm using Contourlet and Discrete Cosine transforms to perceptually shape the watermark by taking care of two conflicting parameters imperceptibility and robustness. Multi Objective Genetic Algorithm is used to decide the optimal watermarking strength (a). Contourlet transform is used to get better quality by taking care of Human Visual System and better robustness under various kinds of linear and non-linear filtering attacks. Discrete Cosine Transform with Zigzag scanning is used to detect the watermark blindly. The watermark embedding process is applied over the sub-band coefficients that lie on the edges and around the edges where distortions are less noticeable. The proposed algorithm provides an excellent tradeoff between robustness and imperceptibility and it is image adaptive. The experimental results show that the proposed algorithm improved the resistance to attacks especially JPEG compression with 5% quality, Median filter with a mask of 6 × 6, resizing up to 20% and various non-linear filtering attacks. Motivation of presented work is to adapt a watermark sequence to those frequency components of an image which provide a transparent and robust watermark.