Research On Digital Watermarking Method Based On3D Mesh Segmentation

The rapid development of computer networks has greatly improved people’s workingefficiency and living standard. It is convenient for people to download and use a variety ofapplications. Information transmission and exchange is also more efficient. The behavior ofillegal copying and dissemination for digital products is emerging in an endless stream. In orderto solve the problem of copyright protection and information security of the digital products,many researchers in academia attach great attention to it. Digital watermarking is a kind ofeffective method to solve the problem which is mentioned above.The three-dimensional model has been widely used in many fields, such as the game industry,animation, computer aided design and also virtual reality. As we all know, the3D model is aform of digital products, but it is essential for people to research on copyright protection of3Dgeometric models. The3D model of digital watermarking technique provides an efficient wayfor copyright protection of digital products.The3D model of digital watermarking technique embeds the watermark information into the3D geometric model to prove that the author possesses the copyright. Considering that thewatermarking techniques of the two-dimensional images cannot be directly applied to thethree-dimensional model, thousands of researchers launched a large number of studies on the3Dgeometric models and also achieved many findings. The watermark information is hidden intothe3D model without perceived while cannot affect the usage value of the carrier.The main contents of this paper contain:1. This paper analyses the background and significance of the3D model of digitalwatermarking technology. It gives an elaboration on the current research and existingmethods about the3D model of digital watermarking and describes the representation andstorage of the3D mesh model. This article discusses the concept of mesh segmentationand presents the evaluation of segmentation algorithm. Besides, the paper provides a briefintroduction on attack analyses and technical difficulties about the3D model of digitalwatermarking.2. In this paper, we proposed a digital watermarking algorithm based on mesh segmentation.The basic approaches are as follows: At the beginning, the3D mesh model is to achievethe invariance to translation, rotation and scaling after the affine transformation. Secondly,we pick up h sub-meshes according to the results of the mesh segmentation procedure. Wecalculate the frequency coefficient matrix for each sub-mesh as well as embed thewatermark information into it. Thirdly, we compare the extracted watermark with theoriginal watermark to evaluate the efficiency of the algorithm we have put forward. Thistechnique possesses the robustness against common mesh attacks such as random noise,meshes cropping, mesh simplification and combined attack. 3. An improved non-blind watermarking method has been put forward according to thedifferent methods on selecting the watermark carrier. First of all, the scheme made theutmost of the results of the mesh segmentation as well as selected h sub-meshes whichcontain a large number of vertices. Secondly, the Cartesian coordinates of the vertices aretransformed to the spherical coordinate system. We select a component to use DWTtransformation and embed the watermark into the frequency domain matrix. The approachis not only robust to random noise, but also robust to mesh simplification. At the sametime, it can efficiently maintain a good visual effect of the3D model.