Robust transparent image watermarking with Shearlet transform and bidiagonal singular value decomposition

Watermarking, means hiding data in digital medium such as image, is a good solution for copyright protection and authentication. Watermarking methods must have a good resistance again various attacks. Wavelet based techniques, due to their compatibility with the human visual system, have been used frequently in this area. In the field of mathematical analysis, several developments of wavelet transform have been proposed. Shearlet transform is a new extension of wavelet that is presented in the past few years. This transform is based on the analysis of multi-resolution and multi-directional and improves presentation of multidimensional data. The main characteristic of Shearlet is good edge presentation of image, so images with different textures can be modeled well using it. In this study, in addition to using Shearlet transform to gain high imperceptibility, bidiagonal singular value decomposition is exploited for robustness and security improvement. Experiments are carried out by various attacks on a wide range of host images with distinct texture characteristics. Results show that the proposed method has good transparency and high robustness against variety types of attacks. Hence, this method can be used in applications that other wavelet-based schemes are not effective enough.