Multiresolution analysis of 3D triangular meshes and application to compression

Sébastien Valette and Rémy Prost


Purpose and context

This work aims at computing multiresolution analysis on 3D triangular meshes [1]

Figure 1 : Multiresolution hierarchy build from the original mesh (left). This representation enables efficient progressive transmission.


Based on the inversion of an irregular subdivision scheme [2], our algorithm is able to construct a complete hierarchy of meshes from one original mesh (see figure 1). This simplification is reversible, and the information to reconstruct the original mesh connectivity from the lowest resolution mesh is a compact code, well suited for connectivity compression. The subdivision paradigm allows our approach to apply the wavelet decomposition on the mesh geometry (the vertices coordinates), the filter bank being constructed by applying the lifting scheme on the “lazy” wavelets filter bank (see figure 2).

Figure 2 : a “lazy” wavelet (left) and its lifted version (right)

Using this wavelet scheme, two compression approaches were proposed : compression with progressive resolution [3] (the mesh connectivity is refined during transmission) or compression with progressive precision [4] (the mesh geometry is refined during transmission).

Download Executable

Click here to download the executable

the source code can be downloaded here.

This programm is distributed under the terms of the GPL license.

copyright CNRS, INSA-Lyon, UCBL, INSERM.


  • This program uses the Visualization ToolKit (VTK). You do not need to download the library, it is already included in the executables)
  • This program uses the range encoder from Michael Schindler for file encoding.