OpenReco
abhibagul.github.io
- Free plan
- Open source
- Windows
- macOS
- Linux
- Web
- API
- Self-hosted
What OpenReco is
OpenReco turns drone and camera imagery into point clouds, textured meshes, orthomosaics, and elevation models. Its node-based workspace can run a complete reconstruction graph or selected stages, with cached results that can be reproduced. The software uses COLMAP CUDA for GPU multi-view stereo and says it can switch automatically to CPU processing. Models can be georeferenced from EXIF GPS data or ground control points. Analysis tools measure volume, profile, area, and distance, and can create vegetation maps using NDVI, ExG, and VARI. Listed exports include LAS, LAZ, PLY, OBJ, GeoTIFF, GeoJSON, DXF, CSV, 3D Tiles, and contour SHP. OpenReco is MIT-licensed open-source desktop software for Windows, macOS, and Linux. The site lists the free plan as free and says no account or subscription is needed to download it. Its named target fields include surveying and mapping, agriculture and forestry, VFX and heritage, construction and earthworks, and robotics and research.
Who it suits
OpenReco may suit people working in surveying, mapping, agriculture, forestry, construction, earthworks, VFX, heritage, robotics, or research. It is aimed at users who need to derive mapped 3D or elevation outputs from drone or camera imagery.
Good to know, and what to watch
- Creates point clouds, textured meshes, and orthomosaics.
- Supports GPS and ground-control-point georeferencing.
- Includes measurement and vegetation-map tools.
- Exports GIS, 3D, and tabular formats.
- Free desktop downloads for Windows, macOS, and Linux.
- Automatic CPU fallback is stated for reconstruction.
- The listed deployment is desktop only.
Plans
Every plan and priceFree
Free plan
The verdict
OpenReco combines image reconstruction, georeferencing, measurement, vegetation analysis, and a broad set of export formats. Its no-account, no-subscription download terms and support for three desktop operating systems are clearly stated.
What the maker publishes
- Account and subscription
- The site says users can download OpenReco with no account and no subscription. source
- Analysis
- Features include volume, profile, area, and distance measurements, plus NDVI, ExG, and VARI vegetation maps. source
- CLI
- The command line includes run, resume, diff, report, stages, batch, export, ui, volume and profile commands. source
- Dense reconstruction
- Dense multi-view stereo depth maps are fused into a high-density point cloud, using COLMAP CUDA with CPU fallback. source
- Desktop UI
- The UI provides a layer tree, schema-driven parameter panels, live run progress and a three.js 3D viewport. source
- Documentation depth
- The repository contains documentation covering discovery, product planning, architecture, pipeline specification and roadmap, plus a detailed README. source
- Download platforms
- The site lists Windows, macOS, and Linux downloads. source
- Elevation products
- The product can rasterize digital surface models and generate contours from elevation data. source
- Export formats
- Listed outputs include LAS, LAZ, PLY, OBJ, GeoTIFF, GeoJSON, DXF, CSV, 3D Tiles and contour shapefiles. source
- Exports
- Listed output formats include LAS, LAZ, PLY, OBJ, GeoTIFF, GeoJSON, DXF, CSV, 3D Tiles, and contour SHP. source
- Georeferencing
- Models can be georeferenced using EXIF GPS or ground control points. source
- Georeferencing
- Models can be positioned using EXIF GPS or ground-control points in an EPSG coordinate reference system. source
- GPU requirements
- NVIDIA dense processing requires the machine's NVIDIA driver and CUDA COLMAP; macOS uses CPU because it has no CUDA. source
- Install options
- The README describes standalone PyInstaller binaries plus installation from Python with pip and an optional bootstrap command. source
- Licence
- OpenReco is released under the MIT licence and its stated dependencies are restricted to permissive BSD, MIT, Apache or similar licences. source
- License
- The site describes OpenReco as MIT licensed and open source. source
- Limit: approximate outputs
- Orthophoto processing is described as approximate and the DTM is based on morphological DSM processing. source
- Limit: Gaussian splatting
- The Gaussian Splatting branch requires torch, gsplat, CUDA, nvcc, CUDA headers and a host C++ compiler, and is not validated on the developer's Windows machine. source
- Limit: maturity
- The README labels the project an incomplete v0.1.1 release and describes it as early even though the UAV pipeline is functional. source
- Limit: missing capabilities
- Multi-image texture blending, PBR, cloud collaboration, learned matching, USD/COPC/3D-Tiles streaming and full GUI collaboration are listed as not yet built or roadmap work. source
- Maker
- The GitHub account identifies the maker as Abhishek Bagul, a full-stack web developer based in India. source
- Maker background
- The maker's profile says he previously worked as an aerial surveyor for three years. source
- Measurements
- Users can calculate cut/fill volume, elevation profiles, areas and distances directly on the model. source
- Mesh generation
- OpenReco reconstructs polygonal surfaces with Poisson meshing and can produce models with millions of vertices. source
Checked against the maker's pages on 15 Oct 2026.