Motion-based MESH clustering for MCDWT compression of 3D animated meshes

We present an efficient compression algorithm for sequences of triangular meshes with fixed connectivity. Our two main contributions in this paper are : a clustering technique which regroups the vertices following the same affine motion; a geometry compensation technique based on the estimation of 3D motion parameters which are modeled by affine transform matrices. Then, a scan-based temporal discrete wavelet transform is applied on the compensated sequence, and the resulting wavelet coefficients are finally encoded by an efficient coding scheme which includes a bit allocation process. Simulation results show that our compression method provides good compression performances compared to some state of the art coders.