ML Anthology

Segmenting Unstructured 3D Points into Surfaces

Pascal Fua, Peter T. Sander
ECCV 1992 pp. 676-680
doi:10.1007/3-540-55426-2_73 /eccv/1992/fua1992eccv-segmenting/

Abstract

We propose an approach for building surfaces from an unsegmented set of 3D points. Local surface patches are estimated and their differential properties are used iteratively to smooth the points while eliminating spurious data, and to group them into more global surfaces. We present results on complex natural scenes using stereo data as our source of 3D information.

PDF ECCV Semantic Scholar

Cite

Text

Fua and Sander. "Segmenting Unstructured 3D Points into Surfaces." European Conference on Computer Vision, 1992. doi:10.1007/3-540-55426-2_73

Markdown

[Fua and Sander. "Segmenting Unstructured 3D Points into Surfaces." European Conference on Computer Vision, 1992.](https://mlanthology.org/eccv/1992/fua1992eccv-segmenting/) doi:10.1007/3-540-55426-2_73

BibTeX

@inproceedings{fua1992eccv-segmenting,
  title     = {{Segmenting Unstructured 3D Points into Surfaces}},
  author    = {Fua, Pascal and Sander, Peter T.},
  booktitle = {European Conference on Computer Vision},
  year      = {1992},
  pages     = {676-680},
  doi       = {10.1007/3-540-55426-2_73},
  url       = {https://mlanthology.org/eccv/1992/fua1992eccv-segmenting/}
}