A Comparison and Evaluation of Multi-View Stereo Reconstruction Algorithms

TLDR

The lack of calibrated multi‑view image datasets with known 3D ground truth has prevented direct comparisons of multi‑view stereo reconstruction algorithms. This paper quantitatively compares several multi‑view stereo reconstruction algorithms. The authors survey multi‑view stereo algorithms, introduce a taxonomy of key properties, and describe a process for acquiring calibrated datasets with high‑accuracy ground truth, an evaluation methodology, and online resources for dataset access and model submission. The results of the quantitative comparison on six benchmark datasets show how state‑of‑the‑art algorithms perform relative to one another.

Abstract

This paper presents a quantitative comparison of several multi-view stereo reconstruction algorithms. Until now, the lack of suitable calibrated multi-view image datasets with known ground truth (3D shape models) has prevented such direct comparisons. In this paper, we first survey multi-view stereo algorithms and compare them qualitatively using a taxonomy that differentiates their key properties. We then describe our process for acquiring and calibrating multiview image datasets with high-accuracy ground truth and introduce our evaluation methodology. Finally, we present the results of our quantitative comparison of state-of-the-art multi-view stereo reconstruction algorithms on six benchmark datasets. The datasets, evaluation details, and instructions for submitting new models are available online at http://vision.middlebury.edu/mview.