Hashed Watermarking Algorithm for 3D Authentication Model

A blind and robust hybrid hashing/watermarking method for inserting watermarks into 3D models is presented in this paper. This method makes the watermarks unnoticeable to the human visual system and extremely difficult to detach. To enhance the performance and the security of the suggested watermarking methodology, the watermark is constructed using hashing function based on Arnold transform and singular value decomposition (SVD). The embedding process hides the watermark by modifying the vertex position depending on measuring the valance or degree of faces around the vertices. The vertices are classified based on their mean curvature values to insert the watermark in optimal regions. Moreover, the watermark not making noticeable changes in the original 3D model and resistant watermark to numerous types of attacks such as translation, adding noise, subdividing, scaling, and rotation. The hybridization between hashing and 3D model watermark enables a secure and robust watermarking method that achieves high levels of imperceptibility and robustness.