Steve Seitz • Brian Curless • James Diebel • Daniel Scharstein • Richard Szeliski

This website accompanies our paper

A Comparison and Evaluation of Multi-View Stereo Reconstruction Algorithms,
CVPR 2006, vol. 1, pages 519-526.

The goal of this project is to provide high quality datasets with which to benchmark and evaluate the performance of multi-view stereo reconstruction algorithms. Each dataset is registered with a ground-truth 3D model acquired via a laser scanning process, to be used as a baseline for measuring accuracy and completeness (the ground truth is not distributed).

• Evaluation results
• Datasets
• How to submit your own results

To stay informed about new additions to the evaluation results or other relevant news, you can subscribe to the mailing list mview-announce@cs.washington.edu.

 

 

---

Other online multiview stereo benchmarks and datasets:

• Robust Vision Challenge
• DTU Robot MVS data set
• ETH3D multiview stereo benchmark
• Tanks and Temples benchmark

 

 

  Support for this work was provided in part by NSF grant IIS-0413169. 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. Support for this work was also provided in part by Microsoft Corporation.