MSVD based image watermarking in NSCT domain

Abstract

Image watermarking in wavelet domain has been emerged as an interesting area of research. However, classical wavelet transform, like DWT, lacks in providing directional information and is shift sensitive. These limitations can be overcome with the use of NSCT (Nonsubsampled contourlet transform), which is shift invariant and provides rich directional information. Therefore, in this work, we propose NSCT based image watermarking using MSVD (Multiresolution singular value decomposition). High directionality of NSCT provides better reconstruction of watermarked images, whereas MSVD provides invariance against geometrical attacks. Experiments have been performed over benchmark datasets. Qualitative and quantitative analysis in terms of PSNR (Peak signal to noise ratio) and CC (Correlation coefficients) shows the effectiveness of the proposed approach.