“…Until now, most of the proposed watermarking algorithms, derived from classical schemes such as [3]- [6], are mainly robust against usual image processing such as low-pass filtering or JPEG compression. However, most of them are still weak against geometrical distortions and malicious attacks [7], [8] (e.g., Stirmark [9], [10], collusion attacks, Dewatermarking Attack [11], etc.). It is quite easy to see that data payload, fidelity, and robustness are often in conflict.…”