3D Mesh Model Digital Watermarking Algorithm

Digital watermarking technology, using the insensitivity of the human senses (visual redundant), and redundancy inherent in the digital signal (ie, redundant data characteristics), embedded people concealed marks in digital multimedia data by signal processing methods. The marks will not be aware of the perception system or re-modified, and also they will not affect the use value of the original content, but production party could recognize and indentify them through the dedicated detectors or reader. Digital watermarking technology could determine if object is protected, monitor the spread of protected and authenticity identification and illegal copying solve the copyright disputes and provide evidence for the court. Therefore, digital watermarking research has important practical significance and application value.In this thesis, the author focuses her studies on information security and copyright maintenance, mading a comprehensive and systematical research in the digital watermarking technology and its application extensive digital watermarking algorithm. Firstly, the author systematically reviews the digital watermarking theories, not only the principle and its development are presented. Secondly, the digital watermarking algorithm, which is the most important component of the research work in the thesis, is also deeply explored and practiced, aiming at the weak points for algorithm to improve. Finally, this paper tests the robustness of watermarking algorithm, focusing on the study of edit attack, noise attack in the new algorithm. The paper based on geometric characteristics of digital watermarking, the main research contents as follows:1. Extracting the feature points:the location determines the watermark embedding watermark robustness and invisibility, so it is crucial to select vertices. There we choose direction changes greatly point as the characteristic points through calculating vertex normals adjacent triangles in every direction.2. Embedding watermarks in the three-dimensional mesh model:the feature points are lined up in descending order, grouped according to the length of the watermark, by changing the coordinates of the feature vertices to embed watermark information.3. Extracting watermark in the mesh models:extract watermarks through the inverse process of embedding them, and judge their correctness. Adopt function Cor to calculate correlation, and then use the correlation to judge whether the watermark is correct.