The role of information theory in watermarking and its application to image watermarking

Abstract

This paper reviews the role of information theory in characterizing the fundamental limits of watermarking systems and in guiding the development of optimal watermark embedding algorithms and optimal attacks. Watermarking can be viewed as a communication problem with side information (in the form of the host signal and/or a cryptographic key) available at the encoder and the decoder. The problem is mathematically defined by distortion constraints, by statistical models for the host signal, and by the information available in the game between the information hider, the attacker, and the decoder. In particular, information theory explains why the performance of watermark decoders that do not have access to the host signal may surprisingly be as good as the performance of decoders that know the host signal. The theory is illustrated with several examples, including an application to image watermarking. Capacity expressions are derived under a parallel-Gaussian model for the host-image source. Sparsity is the single most important property of the source that determines capacity.