A New Blind Detection Method Based on Channel Reliability Estimates for Robust Watermarking

In literature, diversity and attack characterization based blind watermark detectors are proposed for robust watermarking. These detectors, simultaneously, use the reference watermarks to characterize the watermark channel, and the information watermarks to transmit the hidden information. In this paper, a novel blind watermark detection method is presented to improve performance of diversity and attack characterization based blind watermark detection method used in quantization based watermarking system. In the proposed method, first, the reliabilities of the watermark channels are estimated. Then, channel parameters are found by using the first and second order statististics of the channel reliabilities and hence realiable watermark channels are determined by the proposed thresholding method. Finally, the information watermark is recovered by using estimated information watermarks from the reliable watermark channels. The simulation results demonstrate that the proposed blind detection method is more robust than existing methods that are used in the quantization based watermarking system under various channel distortions.