ABSTRACT
M. Miller, G. J. Doerr, and I. J. Cox have proposed a robust high-capacity (RHC) watermarking algorithm through informed coding and embedding. There are two criteria to be considered when we develop the RHC watermarking technique. First, the strategy for embedding and extracting watermarks ensures robustness. Second, the strategy for enlarging the capacity does not affect the robustness of the watermarking. Frequency-domain watermarking possesses strong robustness since the embedded messages are spread out all over the spatial extent of an image. During spatial-frequency transformation, the low frequencies in the transformed domain reflect smooth areas of an image, and the high frequencies reflect the areas with large intensity changes, such as edges and noise. Therefore, due to local spatial similarity the significant coefficients of a transformed image are limited. The chapter illustrates that algorithm can enlarge the watermarking capacity by using the block-based chaotic map and Intersection-based pixels collection approaches.
