A Blind 3D Watermarking Approach for 3D Mesh Using Clustering Based Methods

Citation:
Mona M. Soliman, A. E. Hassanien, and H. M. Ons, "A Blind 3D Watermarking Approach for 3D Mesh Using Clustering Based Methods", IJCVIP - International Journal of Computer Vision and Image Processing, vol. 3, issue 2, pp. 43-53, 2013.

Related External Link

Blind and robust watermarking of 3D mesh aims to embed message into a 3D mesh model such that the mesh is not visually distorted from the original model. An essential condition is that the message should be securely extracted even after the mesh model was processed. This paper explores use of artificial intelligence techniques to build blind and robust 3D-watermarking approach. It is based on clustering 3D vertices into appropriate or inappropriate candidates for watermark insertion using K-means clustering and Self Organization Map (SOM) clustering algorithms. The watermark insertion were performed only on set of selected vertices come out from clustering technique. These vertices are used as candidates for watermark carriers that will hold watermark bits stream. Through the simulations, the authors prove that the proposed approach is robust against various kinds of geometrical attacks such as mesh smoothing, noise addition and mesh cropping.


Tourism