vision.middlebury.edu/stereo/data

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.