Overview

The Muddlebury Stereo Dataset is collected under the support of Middlebury College.The dataset is developed for online evaluation dense two-frame stereo algorithms.

Data Format

2001 Stereo datasets with ground truth

These datasets of piecewise planar scenes were created by Daniel Scharstein, Padma Ugbabe, and Rick Szeliski. Each set contains 9 images (im0.ppm - im8.ppm) and ground-truth disparity maps for images 2 and 6 (disp2.pgm and disp6.pgm). Each ground-truth disparity map is scaled by a factor of 8. For example, a value of 100 in disp2.pgm means that the corresponding pixel in im6.ppm is 12.5 pixels to the left.

2003 Stereo datasets with ground truth

These datasets were created by Daniel Scharstein, Alexander Vandenberg-Rodes, and Rick Szeliski. They consist of high-resolution stereo sequences with complex geometry and pixel-accurate ground-truth disparity data. The ground-truth disparities are acquired using a novel technique that employs structured lighting and does not require the calibration of the light projectors. See our CVPR 2003 paper for more details.

Quarter-size (450 x 375) versions of our new data sets "Cones" and "Teddy" are available for download. Each data set contains 9 color images (im0..im8) and 2 disparity maps (disp2 and disp6). The 9 color images form a multi-baseline stereo sequence, i.e., they are taken from equally-spaced viewpoints along the x-axis from left to right. The images are rectified so that all image motion is purely horizontal. To test a two-view stereo algorithm, the two reference views im2 (left) and im6 (right) should be used. Ground-truth disparites with quarter-pixel accuracy are provided for these two views. Disparities are encoded using a scale factor 4 for gray levels 1 .. 255, while gray level 0 means "unknown disparity". Therefore, the encoded disparity range is 0.25 .. 63.75 pixels.

2005 Stereo datasets with ground truth

These 9 datasets were created by Anna Blasiak, Jeff Wehrwein, and Daniel Scharstein at Middlebury College in the summer of 2005, and were published in conjunction with two CVPR 2007 papers [3, 4]. Each image below links to a directory containing the full-size views and disparity maps. Shown are the left views; moving the mouse over the images shows the right views. We're withholding the true disparity maps for three of the sequences (Computer, Drumsticks, and Dwarves) which we may use in future evaluations.

Art	Books	Dolls
Laundry	Moebius	Reindeer
Computer	Drumsticks	Dwarves

2006 Stereo datasets with ground truth

These 21 datasets were created by Brad Hiebert-Treuer, Sarri Al Nashashibi, and Daniel Scharstein at Middlebury College in the summer of 2006, and were published in conjunction with two CVPR 2007 papers [3, 4]. Each image below links to a directory containing the full-size views and disparity maps. Shown are the left views; moving the mouse over the images shows the right views.

Aloe	Baby1	Baby2

2014 Stereo datasets with ground truth

These 33 datasets were created by Nera Nesic, Porter Westling, Xi Wang, York Kitajima, Greg Krathwohl, and Daniel Scharstein at Middlebury College during 2011-2013, and refined with Heiko Hirschmüller at the DLR Germany during 2014. A detailed description of the acquisition process can be found in GCPR 2014 paper.

20 of the datasets are used in the new Middlebury Stereo Evaluation (10 each for training and test sets). Except for the 10 test datasets, we provide links to directories containing the full-size views and disparity maps. Shown are the left views at 5% resolution; moving the mouse over the images shows the right views.

10 evaluation test sets (GT hidden)
Australia	Bicycle2	Classroom2	Crusade	Djembe
Hoops	Livingroom	Newkuba	Plants	Staircase
10 evaluation training sets with GT
Adirondack: perf, imp	Jadeplant: perf, imp	Motorcycle: perf, imp	Piano: perf, imp	Pipes: perf, imp
Playroom: perf, imp	Playtable: perf, imp	Recycle: perf, imp	Shelves: perf, imp	Vintage: perf, imp
13 additional datasets with GT
Backpack: perf, imp	Bicycle1: perf, imp	Cable: perf, imp	Classroom1: perf, imp	Couch: perf, imp
Flowers: perf, imp	Mask: perf, imp	Shopvac: perf, imp	Sticks: perf, imp	Storage: perf, imp
Sword1: perf, imp	Sword2: perf, imp	Umbrella: perf, imp

Citation

Please use the following citation when referencing the dataset:

@inproceedings{inproceedings,
author = {Scharstein, Daniel and Szeliski, Richard and Zabih, Ramin},
year = {2001},
month = {02},
pages = {131-140},
title = {A taxonomy and evaluation of dense two-frame stereo correspondence algorithm},
volume = {47},
isbn = {0-7695-1327-1},
journal = {Int. J. Comput. Vision},
doi = {10.1109/SMBV.2001.988771}
}