ML Anthology

Optimal Motion Estimation from Multiview Normalized Epipolar Constraint

René Vidal, Yi Ma, Shawn Hsu, Shankar Sastry
ICCV 2001 pp. 34-41
doi:10.1109/ICCV.2001.10054 /iccv/2001/vidal2001iccv-optimal/

Abstract

In this paper, we study the structure from motion problem as a constrained nonlinear least squares problem which minimizes the so called reprojection error subject to all constraints among multiple images. By converting this constrained optimization problem to an unconstrained one, we obtain a multiview version of the normalized epipolar constraint of two views. Such a multiview normalized epipolar constraint serves as a statistically optimal objective function for motion (and structure) estimation. Since such a function is defined naturally on a product of Stiefel manifolds, we show how to use geometric optimization techniques to minimize it. We present experimental results on real images to evaluate the proposed algorithm.

ICCV Semantic Scholar

Cite

Text

Vidal et al. "Optimal Motion Estimation from Multiview Normalized Epipolar Constraint." IEEE/CVF International Conference on Computer Vision, 2001. doi:10.1109/ICCV.2001.10054

Markdown

[Vidal et al. "Optimal Motion Estimation from Multiview Normalized Epipolar Constraint." IEEE/CVF International Conference on Computer Vision, 2001.](https://mlanthology.org/iccv/2001/vidal2001iccv-optimal/) doi:10.1109/ICCV.2001.10054

BibTeX

@inproceedings{vidal2001iccv-optimal,
  title     = {{Optimal Motion Estimation from Multiview Normalized Epipolar Constraint}},
  author    = {Vidal, René and Ma, Yi and Hsu, Shawn and Sastry, Shankar},
  booktitle = {IEEE/CVF International Conference on Computer Vision},
  year      = {2001},
  pages     = {34-41},
  doi       = {10.1109/ICCV.2001.10054},
  url       = {https://mlanthology.org/iccv/2001/vidal2001iccv-optimal/}
}