Rao based watermark detector through two-level hierarchical prior in transform domain

In this work, we propose a transform based blind zero-bit watermark detector, designed based on a hierarchical, two-level image prior. This model is applied through Rao hypothesis test, where we detect the hidden information with unknown amplitude. The proposed method is suitable for wavelet domain watermark detection without the need to estimate the appropriate parameters under the alternative hypothesis. The proposed system shows efficient robustness against a variety of known attacks. The experimental results show that the proposed detector has a good performance with or without attacks in terms of ROC (receiver operating characteristics) curves compared with other known state-of-the-art statistical detectors.