White Rose University Consortium logo
University of Leeds logo University of Sheffield logo York University logo

Robust and Blind 3D Watermarking

Luo, Ming (2006) Robust and Blind 3D Watermarking. PhD thesis, University of York.

[img] Text
Available under License Creative Commons Attribution-Noncommercial-No Derivative Works 2.0 UK: England & Wales.

Download (7Mb)


3D watermarking is a technique to hide some information into the 3D graphical model in such a way that the watermarked object is visually indistinguishable from the original one. A robust and blind 3D watermarking method should be able to detect the embedded message after a certain level of malicious attack without having the original model. 3D watermarking has a great potential of usage in the real world and it can be applied in the copyright protection, database management, graphics authentication and data transmission etc. This thesis proposes four novel robust and blind 3D watermarking methods based on spectral domain and spatial domain. Chapter 2 comprehensively surveys the related literature in the fields of transformed domain methods, spatial domain methods and the watermarking metrics. Chapter 3 proposes a novel 3D watermarking methodology in the spectral domain. The mesh object is decomposed into a set of spectral coefficients which represent the energy of the mesh in different scales. The message is embedded by introducing constraints into the distributions of spectral coefficients. Chapter 4 employs the geodesic distance to carry the bits based on the observation that the distribution of geodesic distance within a range is close to uniform. Two ways of embedding scheme are introduced. One is to modify the mean value and the distribution and the other is to change the variance. A novel Vertex Placement Scheme (VPS) is proposed to move the vertex in order to satisfy the watermarked geodesic distance, without causing significant distortion to the object. Chapter 5 introduces two spatial domain methods which embed the message by changing the distribution of the vertex norms, i.e. the distance from vertex to the object centre. Two methods employ the same histogram mapping function as described in chapter 4. The first method minimizes the surface distortion by selecting a candidate point over the neighbourhood which introduces the minimum error. The second method employs the Levenberg-Marquardt optimization method to find the best possible solution to ensure that the surface distortion is truly minimum with respect to a novel surface error function. The algorithms proposed in this thesis significantly improve the visual quality of the watermarked object while the watermark detection robustness is at a relatively high level. The robustness of the proposed methods is increasing from the methods presented in Chapter 3 to Chapter 5 while the surface distortion is decreasing for these methods. The second algorithm proposed in Chapter 5 achieves the best overall performance in the aspect of visual quality and robustness. In Chapter 6, we conclude the thesis by addressing the weakness and propose potential future research work.

Item Type: Thesis (PhD)
Keywords: Watermarking, surface error, spectral analysis, geodesic distance, histogram mapping
Academic Units: The University of York > Computer Science (York)
Identification Number/EthosID: uk.bl.ethos.516402
Depositing User: Mr Ming Luo
Date Deposited: 24 May 2010 11:06
Last Modified: 08 Sep 2016 12:15
URI: http://etheses.whiterose.ac.uk/id/eprint/863

You do not need to contact us to get a copy of this thesis. Please use the 'Download' link(s) above to get a copy.
You can contact us about this thesis. If you need to make a general enquiry, please see the Contact us page.

Actions (repository staff only: login required)