If you have any problems related to the accessibility of any content (or if you want to request that a specific publication be accessible), please contact us at email@example.com.
Third Generation 3D Watermarking: Applied Computational Intelligence Techniques
AuthorMotwani, Mukesh C.
AdvisorHarris, Frederick C.
Computer Science and Engineering
AltmetricsView Usage Statistics
With the explosion of multimedia content and its ease of distribution over Internet, there is a need for copyright protection of digital content. Whether it is music albums swapped over peer to peer networks or video files uploaded over websites such as YouTube.com or 3D models such as Shrek, artists need to protect their ownership of content. 3D Watermarking provides a deterrent to piracy of 3D models by embedding a hidden piece of information in the original content. Watermarking algorithms have a basic requirement that the watermark should be imperceptible to avoid being detected and not cause visible distortion to the viewer. It is desired that the amount of watermark inserted at each location in the 3D model should be as high as possible to withstand intentional attacks and to allow insertion of multiple or redundant or biometric watermarks. Insertion of watermark at as many random locations in the 3D model as possible will make it extremely difficult for an attacker to find the watermark and then make substantial changes in the 3D model to remove or overwrite the watermark. However, inserting large amounts of information as watermark can cause distortion. The watermark should also be robust to withstand unintentional attacks. The design of watermarking algorithms involves a trade off between imperceptibility, amount of information inserted and robustness. The first generation of 3D watermarking techniques inserted low amount of watermark based on spatial geometry and have poor robustness. The second generation of algorithms explored use of multiresolution transform for watermark insertion to improve the robustness. This dissertation explores use of computational intelligence techniques to build third generation watermarking algorithms, that insert robust, high amount of watermark and go the extra mile in terms of hiding more information than the first and second generation techniques. The focus of this study is to optimize the energy of the watermark and intelligently selecting information pockets in 3D model for watermark insertion and at the same time still maintaining randomness in the process to avoid detection. Use of Fuzzy Logic, Genetic Algorithms, and Artificial Neural Networks are proposed and assessed.