ML Anthology

On Exploiting Occlusions in Multiple-View Geometry

Paolo Favaro, Alessandro Duci, Yi Ma, Stefano Soatto
ICCV 2003 pp. 479-486
doi:10.1109/ICCV.2003.1238386 /iccv/2003/favaro2003iccv-exploiting/

Abstract

Occlusions are commonplace in man-made and natu-ral environments; they often result in photometric features where a line terminates at an occluding boundary, resem-bling a “T”. We show that the 2-D motion of such T-junctions in multiple views carries non-trivial information on the 3-D structure of the scene and its motion relative to the camera. We show how the constraint among multiple views of T-junctions can be used to reliably detect them and differentiate them from ordinary point features. Finally, we propose an integrated algorithm to recursively and causally estimate structure and motion in the presence of T-junctions along with other point-features. 1

PDF ICCV Semantic Scholar

Cite

Text

Favaro et al. "On Exploiting Occlusions in Multiple-View Geometry." IEEE/CVF International Conference on Computer Vision, 2003. doi:10.1109/ICCV.2003.1238386

Markdown

[Favaro et al. "On Exploiting Occlusions in Multiple-View Geometry." IEEE/CVF International Conference on Computer Vision, 2003.](https://mlanthology.org/iccv/2003/favaro2003iccv-exploiting/) doi:10.1109/ICCV.2003.1238386

BibTeX

@inproceedings{favaro2003iccv-exploiting,
  title     = {{On Exploiting Occlusions in Multiple-View Geometry}},
  author    = {Favaro, Paolo and Duci, Alessandro and Ma, Yi and Soatto, Stefano},
  booktitle = {IEEE/CVF International Conference on Computer Vision},
  year      = {2003},
  pages     = {479-486},
  doi       = {10.1109/ICCV.2003.1238386},
  url       = {https://mlanthology.org/iccv/2003/favaro2003iccv-exploiting/}
}