Robust 3D Watermarking Using Vertex Smoothness Measure

Watermarking of 3D data has gathered renewed interest due to the explosion of graphic content and animated motion movies. This paper discusses non-blind watermarking of 3D models in the spatial domain. The proposed method estimates the local smoothness variation of the mesh to select vertices for inserting a watermark. Smoothness variation of the surface represented by the 1ring neighborhood of each vertex is computed by the average angle difference between the surface normal and the average normal. Each vertex of the mesh is labeled in one of the bins corresponding to varying degrees of local smoothness variation from the low to the moderate to the highest variation. Vertices with the label of moderate local smoothness variation are then selected for the insertion of a random watermark. Simulation results prove that the inserted watermark is robust against cropping, affine operations and noise attacks, and at the same time imperceptibility is maintained.