openMVG/openMVG

Health v2 · maintenance only

Adoption not part of the score

6542 stars · 1719 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

OpenMVG is a C++ library for multiple view geometry and Structure from Motion (SfM), providing end-to-end 3D reconstruction from images. It ships as libraries, command-line binaries, and pipelines covering feature detection, matching, camera models, robust estimation, and scene reconstruction.

Use cases

• reconstruct 3d scenes from photo collections
• run structure from motion on drone imagery
• estimate camera poses and triangulate 3d points from multiple views
• detect and match image features for photogrammetry
• export sparse reconstructions for dense mesh generation
• prototype multiple view geometry research algorithms

When to choose

• you need a readable, well-tested C++ SfM pipeline you can extend
• you want sparse 3D reconstruction from unordered image sets
• you need cross-platform (Linux/macOS/Windows/Android/iOS) computer vision geometry building blocks
• you want to feed reconstructions into MVS tools for dense point clouds

When to avoid

• you need dense reconstruction or textured meshes out of the box - pair it with an MVS tool like OpenMVS
• you want a GUI photogrammetry application rather than libraries and CLI binaries
• you need real-time SLAM rather than offline SfM
• you prefer Python-first tooling

Facets

library · maturity active

computer-vision image-processing graphics sdk cli computer-vision graphics windows cpp cross-platform structure-from-motion 3d-reconstruction photogrammetry multiple-view-geometry feature-matching camera-models algorithms linux macos android ios

Member repositories

For agents

markdown · JSON · MCP: product_card(name="openMVG/openMVG")