COLMAP
| Original author(s) | Johannes L. Schönberger |
|---|---|
| Initial release | April 4, 2016[1] |
| Stable release | 3.12.0
/ June 30, 2025[2] |
| Written in | C++ |
| Engine | |
| Operating system | Windows, Linux, macOS[3] |
| Platform | Cross-platform |
| Type | Photogrammetry, 3D reconstruction |
| License | BSD 3-clause[3] |
| Website | colmap |
Search COLMAP on Amazon.
COLMAP is a free and open-source software for photogrammetry and 3D reconstruction from images. It provides a general-purpose pipeline for structure from motion (SfM) and multi-view stereo (MVS), enabling the creation of 3D models from photographs.[3] COLMAP was developed by computer vision researcher Johannes L. Schönberger and was first released in 2016.[1] The software is cross-platform, available for Windows, Linux, and macOS.[3] It is released under the BSD license.[3]
History
COLMAP originated from research at ETH Zurich by Johannes L. Schönberger and colleagues. Its core methods were introduced in two 2016 publications, Structure-from-Motion Revisited and Pixelwise View Selection for Unstructured Multi-View Stereo, presented at the IEEE CVPR and ECCV, respectively.[4][5] These papers presented improvements to SfM and MVS techniques and released the algorithms as open-source software. The first public release of COLMAP was on 4 April 2016.[1]
Features
COLMAP provides an end-to-end pipeline for photogrammetric 3D reconstruction from images. Its functionality is divided into:
- Sparse reconstruction (SfM): Detects and matches image features (e.g., using SIFT), estimates camera parameters, and reconstructs a sparse 3D point cloud using incremental bundle adjustment.[4] COLMAP uses the Ceres Solver for non-linear optimization.
- Dense reconstruction (MVS): Computes depth maps for input images with pixelwise multi-view selection, then fuses them to create a dense point cloud.[5]
- User interface: Available both as a graphical user interface (GUI) and a command-line tool.
COLMAP outputs camera parameters, sparse/dense point clouds, and mesh representations in standard formats.
Reception and impact
Since its initial release, COLMAP has been used in in academic research and practical projects.[6]
See also
References
- ↑ 1.0 1.1 1.2 "Changelog – COLMAP 3.12 documentation". Retrieved 2024-06-15.
COLMAP 1.0 (4/4/2016) – Initial release of COLMAP.
- ↑ Cite error: Invalid
<ref>tag; no text was provided for refs namedlatest_release - ↑ 3.0 3.1 3.2 3.3 3.4 Schönberger, Johannes L. "COLMAP - Structure-from-Motion and Multi-View Stereo". Retrieved 2024-06-15.
The software runs under Windows, Linux and Mac... COLMAP is licensed under the BSD License.
- ↑ 4.0 4.1 Schönberger, Johannes L.; Frahm, Jan-Michael (2016). Structure-from-Motion Revisited. CVPR.
- ↑ 5.0 5.1 Schönberger, Johannes L.; Zheng, Enliang; Frahm, Jan-Michael; Pollefeys, Marc (2016). "Pixelwise View Selection for Unstructured Multi-View Stereo". In Leibe, Bastian; Matas, Jiri; Sebe, Nicu; Welling, Max. Computer Vision – ECCV 2016. Lecture Notes in Computer Science. 9907. Cham: Springer International Publishing. pp. 501–518. doi:10.1007/978-3-319-46487-9_31. ISBN 978-3-319-46487-9. Search this book on
- ↑ Stathopoulou, E. K.; Franken, J.; Remondino, F. (2018). "Open-Source Image-Based 3D Reconstruction Pipelines: Review, Comparison and Evaluation". Photogrammetric Record. 33 (161): 27–47. doi:10.1111/phor.12219.
This article "COLMAP" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:COLMAP. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
