Date of Award
Doctor of Philosophy in Electrical Engineering - (Ph.D.)
Electrical and Computer Engineering
Yun Q. Shi
Constantine N. Manikopoulos
Frank Y. Shih
Edward Kin-Ming Wong
Image data hiding represents a class of processes used to embed data into cover images.
Robustness is one of the basic requirements for image data hiding. In the first part of this dissertation, 2D and 3D interleaving techniques associated with error-correction-code (ECC) are proposed to significantly improve the robustness of hidden data against burst errors.
In most cases, the cover image cannot be inverted back to the original image after the hidden data are retrieved. In this dissertation, one novel reversible (lossless) data hiding technique is then introduced. This technique is based on the histogram modification, which can embed a large amount of data while keeping a very high visual quality for all images. The performance is hence better than most existing reversible data hiding algorithms.
However, most of the existing lossless data hiding algorithms are fragile in the sense that the hidden data cannot be extracted correctly after compression or small alteration. In the last part of this dissertation, we then propose a novel robust lossless data hiding technique based on patchwork idea and spatial domain pixel modification. This technique does not generate annoying salt-pepper noise at all, which is unavoidable in the other existing robust lossless data hiding algorithm. This technique has been successfully applied to many commonly used images, thus demonstrating its generality.
Ni, Zhicheng, "Image data hiding" (2005). Dissertations. 675.