AliceVision
Open-source 3D Computer Vision

AliceVisionTurn images into 3D.

AliceVision is a 3D Computer Vision framework for 3D reconstruction, camera tracking and HDR panoramas. Meshroom is its node-based desktop app — start from a ready-made pipeline, or wire your own.

Download Meshroom Make a Donation
Windows · LinuxLicense MPL-2.0 ↗meshroom.org ↗
A carved head on a table, half of it shown as the reconstructed 3D mesh in blue
3D reconstructionImages to textured mesh
Camera trackingSolve shots for VFX
HDR panoramaBrackets, fisheye, 360°
Photometric stereoNormals from lighting
RAW conversionColor-managed images
Ready-made pipelines

A template for every job.#

Each template is a Meshroom graph built from AliceVision nodes. Open one, drop in your images, press Start — or fork it into your own pipeline.

Browse all templates on GitHub →

Photogrammetry

Scenes, single objects, turntables and two-sided captures — from raw photos to a textured mesh.

photogrammetryphotogrammetryDraftphotogrammetryObjectphotogrammetryObjectTurntable

Camera tracking

Matchmove from plates, with or without lens calibration, nodal shots, depth-assisted variants and LiDAR.

cameraTrackingcameraTrackingWithoutCalibrationnodalCameraTrackingcameraTrackingDepth

HDR & panorama

Merge bracketed LDR exposures into HDR, then stitch panoramas from perspective or fisheye images.

hdrFusionpanoramaHdrpanoramaFisheyeHdr

Photometric stereo

Recover fine surface normals from a fixed camera and varying light — single view or multi-view.

photometricStereomulti-viewPhotometricStereo

Calibration & utilities

Lens distortion from checkerboards, color charts, RAW development and LiDAR scan meshing.

distortionCalibrationcolorCalibrationrawImageConversion

Build your own

Combine AliceVision nodes — from FeatureExtraction to ExportUSD — and save any graph as a default template.

Read the node reference →

Inside the photogrammetry template#

  1. 01 · InputRead camerasImage metadata and sensor database.CameraInit
  2. 02 · FeaturesFind & matchKeypoints shared between images.FeatureExtraction · ImageMatching · FeatureMatching
  3. 03 · SfMSolve camerasPoses and a sparse point cloud.TracksBuilding · SfMBootStrapping · SfMExpanding
  4. 04 · MVSDense depthA depth map per view, on the GPU.DepthMap · DepthMapFilter
  5. 05 · OutputMesh & textureMesh with UV textures.Meshing · MeshFiltering · Texturing
Two renders of a reconstructed carved lion head
Meshroom showing the image gallery, a source image and the solved cameras around a sparse point cloud
Meshroom

A node graph you can see into.#

  • GPU-accelerated, scales to large image sets
  • Every node cached — change a parameter, recompute only what depends on it
  • Retexture after retopology in Blender, ZBrush or Maya
Meshroom HDR panorama workflow with fisheye images
HDR panoramasFuse brackets, estimate the fisheye circle, use motorized-head files.
Meshroom submitting a graph to a render farm
Render farm submissionPer-node CPU, RAM and GPU rules pick the right machines.
A custom Meshroom graph with branching mesh outputs
Custom pipelinesBranch, compare and save your graph as the default template.

Wherever images meet 3D#

Film and VFX, games, science, cultural heritage, architecture, engineering, robotics and more — a few examples below.

See results on Sketchfab →
Illustration of a film set scanned with cameras
Cinema & VFXSet and prop scanning, matchmove
Illustration of scientists digitizing insect specimens
ScienceSpecimen digitization
Illustration of archaeologists capturing ruins
Cultural heritageMonuments and sculpture

Libraries & integrations#

AliceVisionThe C++ 3D Computer Vision framework.library CCTagConcentric-circle markers, sub-pixel detection.library PopSiftSIFT features on the GPU.library BlenderImport for retopology, send back to texture.integration HoudiniReconstruction inside Houdini via SideFX Labs.integration
Learn

Watch and learn Meshroom#

All videos on YouTube →
YouTube playlist · start here

Quick Introduction

From your first images to a textured model, in a few short videos. Open on YouTube ↗

YouTube playlist

Tutorials

Go deeper: specific workflows, nodes and integrations. Open on YouTube ↗

History

Two thousand years in the making.#

From the camera obscura to aerial surveys, and from Snellius to the 5-point algorithm: AliceVision stands on a long line of ideas about measuring the world from images.

Explore the timeline →
  1. 500 BCCamera obscura
  2. 1826Niépce, the first photograph
  3. 1911First aerial photogrammetry

Open source, built between labs and studios.#

State-of-the-art algorithms that can be tested, analyzed and reused — with the robustness production demands.

Sponsors & partners

Epic Games Czech Technical University Simula INP Toulouse

Using AliceVision or Meshroom at work? Become a corporate sponsor of the AliceVision Association.

About

The AliceVision Association#

The AliceVision project is supported by the AliceVision Association, a non-profit organization whose ambition is to democratize 3D digitization technologies from photographs.

The association was created in July 2020 by 14 founding members from the European academic and industry sector, who have been collaborating in the field of 3D computer vision for cultural and creative industries since 2010. It considers 3D digitization technologies essential for many applications for the common good, especially for culture and artistic creation.

Computer vision researcher, developer, 3D scanning or photogrammetry expert? You can contribute to the association and help expand the community.

Contact us

Public mailing listDiscuss with the communityforum@alicevision.org The teamContact us privatelyteam@alicevision.org Association boardContact the board privatelyassociation@alicevision.org