cdcseacave/openMVS

Health v2 · maintenance only

Adoption not part of the score

4100 stars · 984 forks observed · 2026-08-28

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

OpenMVS is an open-source C++ library for Multi-View Stereo 3D reconstruction, taking camera poses and a sparse point-cloud as input and producing a dense point-cloud, refined mesh, and textured mesh. It complements Structure-from-Motion pipelines like OpenMVG by covering the final stage of the photogrammetry workflow.

Use cases

• reconstruct a dense 3D point cloud from a set of posed images
• generate a textured mesh from a sparse point cloud and camera poses
• refine a 3D mesh to recover fine surface details
• build the texturing stage of a photogrammetry pipeline
• pair with OpenMVG to go from photos to a full 3D model
• research in multi-view stereo and stereo matching algorithms

When to choose

• you already have camera poses and a sparse point cloud from SfM and need dense reconstruction
• you need an open-source, scriptable MVS pipeline with CLI tools
• you want full control over dense point cloud, meshing, refinement, and texturing stages
• you are doing computer-vision research on multi-view stereo

When to avoid

• you need an end-to-end photo-to-model tool with a GUI
• you only have raw images without camera poses or a sparse point cloud
• you need a permissively licensed library (it is AGPL-3.0)
• you want real-time or lightweight 3D scanning on mobile hardware

Facets

library · maturity active

computer-vision image-processing graphics computer-vision graphics photography windows cpp multi-view-stereo 3d-reconstruction photogrammetry dense-point-cloud mesh-reconstruction mesh-refinement mesh-texturing structure-from-motion patch-match semi-global-matching linux macos

Member repositories

For agents

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