ML Anthology

Evaluation of Cost Functions for Stereo Matching

Heiko Hirschmüller, Daniel Scharstein
CVPR 2007
doi:10.1109/CVPR.2007.383248 /cvpr/2007/hirschmuller2007cvpr-evaluation/

Abstract

Stereo correspondence methods rely on matching costs for computing the similarity of image locations. In this paper we evaluate the insensitivity of different matching costs with respect to radiometric variations of the input images. We consider both pixel-based and window-based variants and measure their performance in the presence of global intensity changes (e.g., due to gain and exposure differences), local intensity changes (e.g., due to vignetting, non-Lambertian surfaces, and varying lighting), and noise. Using existing stereo datasets with ground-truth disparities as well as six new datasets taken under controlled changes of exposure and lighting, we evaluate the different costs with a local, a semi-global, and a global stereo method.

PDF CVPR Semantic Scholar

Cite

Text

Hirschmüller and Scharstein. "Evaluation of Cost Functions for Stereo Matching." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2007. doi:10.1109/CVPR.2007.383248

Markdown

[Hirschmüller and Scharstein. "Evaluation of Cost Functions for Stereo Matching." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2007.](https://mlanthology.org/cvpr/2007/hirschmuller2007cvpr-evaluation/) doi:10.1109/CVPR.2007.383248

BibTeX

@inproceedings{hirschmuller2007cvpr-evaluation,
  title     = {{Evaluation of Cost Functions for Stereo Matching}},
  author    = {Hirschmüller, Heiko and Scharstein, Daniel},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2007},
  doi       = {10.1109/CVPR.2007.383248},
  url       = {https://mlanthology.org/cvpr/2007/hirschmuller2007cvpr-evaluation/}
}