Digital watermarking is the enabling technology to prove ownership on copyrighted material, detect originators of illegally made copies, monitor the usage of the copyrighted multimedia data, and analyze the spread spectrum of the data over networks and servers. Most watermarking methods for images and video can be viewed as a communications problem in which the watermark must be transmitted and received through a watermark channel. This channel includes distortions resulting from attacks and interference from the original digital data [1]. It is well accepted that an effective watermarking scheme may be described as the secure, imperceptible, robust communication of information by direct embedding in and retrieval from digital data. For verifying the security and robustness of watermarking algorithms, specific attacks have to be applied to test them. In this paper, using a theoretical approach based on random processes, signal processes, and communication theory, we propose a stochast...