A Blind Robust 3D-Watermarking Scheme Based on Progressive Mesh and Self Organization Maps

Most of progressive mesh (PM) transmission techniques, consist in iteratively decimating the mesh, while storing the information necessary to the process inversion. During the transmission or visualization the 3D content can be duplicated and redistributed by a pirate. Digital watermarking is considered as a good solution to this emerging problem. This paper focus on introducing a novel robust and blind mesh watermarking schema by converting the original triangle mesh into a multi-resolution format, consisting of a coarse base mesh and a sequence of refinement, the watermark bits are inserted through progressive mesh level of details, and extracted at refinement stage without any need for the original model. The watermark insertion is performed only on set of marked vertices come out from Self Organization Maps (SOM) clustering neural network. These vertices are used as candidates for watermark carriers that will hold watermark bits through progressive mesh transmission. The robustness of proposed techniques is evaluated experimentally by simulating attacks such as mesh smoothing, noise addition and mesh cropping.