A Histogram Based Watermarking Algorithm Robust to Geometric Distortions

Most of the histogram-based watermarking methods focused on the geometric invariance feature more while paying less attention to the robustness to some common image processing attacks like JPEG compression. In this paper, we propose a new histogram modification scheme considering the visual quality of the watermarked image while combining block dividing and histogram statistics for the first time in watermarking scheme. We use the mean pixel value of divided blocks to calculate histogram and the mean square error to select blocks to be modified based on HVS. We can obtain a good trade-off between robustness to geometric attacks and common image processing attacks by adjusting the block size. Finally the complete embedding and extracting algorithms are given. The experiment shows that the proposed approach has an excellent robustness against geometric attacks and some normal attacks, such as adding noise and JPEG compression.