AliceVision: a photogrammetric computer vision framework for 3D reconstruction and camera tracking

What it solves

AliceVision addresses the challenge of photogrammetry—the process of reversing the projection of a 3D scene onto 2D photographs to infer the original 3D geometry. It provides a framework for 3D reconstruction and camera tracking from a set of unordered photographs or videos.

How it works

It utilizes a pipeline of state-of-the-art computer vision algorithms to process 2D images and reconstruct the 3D scene. While the core engine is a set of algorithms, users can interact with the pipeline through Meshroom, which provides a graphical user interface (GUI) and a command-line interface for launching reconstructions and creating custom Python scripts for automation.

Who it’s for

This framework is designed for researchers and industry professionals who need robust, production-quality 3D reconstruction and camera tracking tools based on cutting-edge computer vision research.

Highlights

• Photogrammetric Framework: Specialized in 3D reconstruction and camera tracking.
• Production-Ready: Developed through collaboration between academia and industry for robustness and quality.
• Flexible Interface: Accessible via Meshroom for GUI-based workflows or Python for custom automation.
• Open Source: Released under the MPLv2 license.