What Is AliceVision and What Can You Do With It?
AliceVision is a photogrammetric computer vision framework for 3D reconstruction and camera tracking. It is a library and framework rather than a one-click application, so it fits developers and advanced users who want to build or script photogrammetry pipelines. If you want the same underlying technology with a graphical interface, Meshroom is the companion application to look at instead.
What AliceVision actually is
The project describes itself as a "Photogrammetric Computer Vision framework for 3D Reconstruction and Camera Tracking." Two words in that description matter:
- Framework / library — it provides building blocks and algorithms you call from your own code or pipeline, not a finished end-user product.
- Photogrammetric — it works from photographs, using the geometry of overlapping images to recover 3D structure and camera positions.
That combination is what separates it from a general-purpose image library. The problem domain is specifically turning photo sets into 3D data and camera motion.
AliceVision vs Meshroom
This is the most common point of confusion, so it is worth stating plainly:
| AliceVision | Meshroom | |
|---|---|---|
| What it is | The underlying framework/library | A GUI application built on it |
| Who it is for | Developers, pipeline builders, advanced users | Users who want a visual, node-based workflow |
| How you use it | From code or scripts | Through a graphical interface |
Use the framework when you need to integrate reconstruction into your own software, automate it, or control individual steps. Use the GUI when you want to run a reconstruction without writing code. They are not competitors — one is the engine, the other is a front end.
Core capabilities
Based on the project's own description, the framework covers:
- 3D reconstruction — building 3D geometry from photographs.
- Camera tracking — recovering camera positions and motion.
- HDR — high dynamic range processing.
- Panorama — panoramic image processing.
The first two are the headline functions; HDR and panorama are supporting capabilities within the same computer vision toolkit.
Typical inputs and outputs
Because this is a photogrammetry framework, the natural input is a set of overlapping photographs of a subject or scene. The expected outputs are the products of reconstruction and tracking: 3D structure and camera information derived from those images.
The exact file formats and intermediate results depend on which parts of the framework you call, so treat the input/output specifics as something to confirm against the current documentation rather than assume.
Who should use it
AliceVision suits you if:
- You are comfortable working with a library or framework rather than a GUI.
- You need to script, automate, or embed photogrammetric reconstruction in a larger pipeline.
- You want control over individual processing stages.
It is a poor fit if you want a casual, one-click tool — in that case, the GUI application is the more appropriate entry point.
Practical next step
Decide first which layer you need. If you want to use photogrammetry, start with the GUI. If you want to build with photogrammetry, the framework is the right layer, and you should check its current documentation for the specific modules, inputs, and outputs relevant to your pipeline.