Middlebury Stereo Datasets
  |
2001 datasets
- 6 datasets of piecewise planar scenes [1] (Sawtooth, Venus, Bull, Poster, Barn1, Barn2) |
  |
2003 datasets
- 2 datasets with ground truth obtained using structured light [2] (Cones, Teddy) |
  |
2005 datasets
- 9 datasets obtained using the technique of [2], published in [3, 4] (Art, Books, Dolls, Laundry, Moebius, Reindeer, Computer, Drumsticks, Dwarves) |
  |
2006 datasets
- 21 datasets obtained using the technique of [2], published in [3, 4] (Aloe, Baby1-3, Bowling1-2, Cloth1-4, Flowerpots, Lampshade1-2, Midd1-2, Monopoly, Plastic, Rocks1-2, Wood1-2) |
  |
2014 datasets - 33 datasets obtained using the technique of [5] |
  |
2021 mobile datasets - 24 datasets obtained with a mobile device on a robot arm, using the technique of [5] |
How to cite our datasets:
We grant permission to use and publish all images and disparity
maps on this website. However, if you use our datasets,
we request that you cite the appropriate
paper(s): [1] for the 2001 datasets, [2] for the 2003 datasets,
[3] or [4] for the 2005 and 2006 datasets, and
[5] for the 2014 and 2021 datasets.
References:
| [1] | D. Scharstein and R. Szeliski.
A taxonomy and evaluation of dense
two-frame stereo correspondence algorithms. International Journal of Computer Vision, 47(1/2/3):7-42, April-June 2002. |
| [2] | D. Scharstein and R. Szeliski.
High-accuracy stereo depth maps using structured light. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2003), volume 1, pages 195-202, Madison, WI, June 2003. |
| [3] | D. Scharstein and C. Pal.
Learning
conditional random fields for stereo. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2007), Minneapolis, MN, June 2007. |
| [4] | H. Hirschmüller and D. Scharstein.
Evaluation
of cost functions for stereo matching. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2007), Minneapolis, MN, June 2007. |
| [5] | D. Scharstein, H. Hirschmüller, Y. Kitajima, G. Krathwohl,
N. Nesic, X. Wang, and P. Westling.
High-resolution stereo datasets with subpixel-accurate
ground truth. In German Conference on Pattern Recognition (GCPR 2014), Münster, Germany, September 2014. |
Support for this work was provided in part by NSF CAREER grant 9984485 and NSF grants IIS-0413169, IIS-0917109, IIS-1320715, and IIS-1718376. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.