A Novel HVS-based Watermarking Scheme in Contourlet Transform Domain

In this paper, a novel watermarking technique in contourlet transform (CT) domain is presented. The proposed algorithm takes advantage of a multiscale framework and multi-directionality to extract the significant frequency, luminance and texture component in an image. Unlike the conventional methods in the contourlet domain, mask function is accomplished pixel by pixel by taking into account the frequency, the luminance and the texture content of all the image subbands including the low-pass subband and directional subbands. The adaptive nature of the novel method allows the scheme to be adaptive in terms of the imperceptibility and robustness. The watermark is detected by computing the correlation. Finally, the experimental results demonstrate the imperceptibility and the robustness against standard watermarking attacks.