Refractive Height Fields from Single and Multiple Images

Abstract

We propose a novel framework for reconstructing homogenous, transparent, refractive height-fields from a single viewpoint. The height-field is imaged against a known planar background, or sequence of backgrounds. Unlike existing approaches that do a point-by-point reconstruction - which is known to have intractable ambiguities - our method estimates and optimizes for the entire height-field at the same time. The formulation supports shape recovery from measured distortions (deflections) or directly from the images themselves, including from a single image. We report results for a variety of refractive height-fields showing significant improvement over prior art.

Cite

Text

Shan et al. "Refractive Height Fields from Single and Multiple Images." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2012. doi:10.1109/CVPR.2012.6247687

Markdown

[Shan et al. "Refractive Height Fields from Single and Multiple Images." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2012.](https://mlanthology.org/cvpr/2012/shan2012cvpr-refractive/) doi:10.1109/CVPR.2012.6247687

BibTeX

@inproceedings{shan2012cvpr-refractive,
  title     = {{Refractive Height Fields from Single and Multiple Images}},
  author    = {Shan, Qi and Agarwal, Sameer and Curless, Brian},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2012},
  pages     = {286-293},
  doi       = {10.1109/CVPR.2012.6247687},
  url       = {https://mlanthology.org/cvpr/2012/shan2012cvpr-refractive/}
}