Title

Three-dimensional digital data embedding and detection

Date of Completion

January 2004

Keywords

Engineering, Electronics and Electrical

Degree

Ph.D.

Abstract

The World Wide Web has become a major channel for trading information in the digital era in which we live. Therefore, means to protect information against illegal copying are crucial. Amongst these is watermarking in which a piece of information is hidden within another such that the quality of the host is not substantially affected. This thesis presents a novel methodology for information watermarking and hiding. The proposed watermarking/hiding algorithm balances between preserving the quality of the host information, securing the hidden information, and protecting the hidden information against removal trials. Also, a novel methodology for optimum distortion tolerant detection is presented. This methodology aims at detecting the recovered hidden image to confirm the authenticity of the host data. This methodology extends the use of the proposed data hiding and detection algorithms to 3D objects. A novel distortion-tolerant maximum likelihood algorithm is developed and utilized for detecting the hidden 2D/3D data. The 3D objects are presented using optically generated holograms. The watermarking process utilizes double phase encoding to increase the security and robustness of the hidden image/3D object against removal trials. ^ The watermarking methodologies and algorithms presented in this thesis are believed to enhance the security of information transferred digitally via the World Wide Web. To the best of the knowledge of the author, this work presents the first attempt to utilize double phase encoding in the watermarking process. A novel aspect of this research is the use of 3D objects in the form of digital hologram instead of a 2D image, both as a host or hidden object. Also, the methodologies presented in this research are capable of hiding more that one image/3D object within the host image/3D object. In addition, distortion tolerant optimum recognition algorithm presented in this work is capable of detecting a distorted 3D object buried in noise accurately. ^ The validity and performance of the proposed watermarking/detection methodologies are verified through mathematical analysis, computer simulations, and extensive experiments. ^