Cells classification data structures for faster isosurface extraction.

in: Proc. 28th Gocad Meeting, Nancy

Abstract

When extracting an isosurface, it has been proved that, on average, only a small subset of the cells of the considered grid is intersected by the isosurface. Thus, pre-classification methods have been developed to pre-select the subset of intersected cells prior to the tesselation stage. In this paper, we first present several well known general classification algorithms like Octrees, Kd-trees, Interval Trees, Seed sets... We also propose a new adaptative algorithm particularly well suited for isosurface extraction. At last, we provide an extensive comparison between most relevant algorithms in the particular case of isosurface extraction. We show that, with our adaptative classification algorithm, the isosurface extraction stage could be accelerated by up to a factor of 5x.

Download / Links

    BibTeX Reference

    @inproceedings{132_buatois,
     abstract = { When extracting an isosurface, it has been proved that, on average, only a small subset of the cells of
    the considered grid is intersected by the isosurface. Thus, pre-classification methods have been developed
    to pre-select the subset of intersected cells prior to the tesselation stage.
    In this paper, we first present several well known general classification algorithms like Octrees, Kd-trees,
    Interval Trees, Seed sets... We also propose a new adaptative algorithm particularly well suited for isosurface
    extraction. At last, we provide an extensive comparison between most relevant algorithms in the particular
    case of isosurface extraction. We show that, with our adaptative classification algorithm, the isosurface extraction
    stage could be accelerated by up to a factor of 5x. },
     author = { Buatois, Luc AND Caumon, Guillaume },
     booktitle = { Proc. 28th Gocad Meeting, Nancy },
     title = { Cells classification data structures for faster isosurface extraction. },
     year = { 2008 }
    }