alicevision/AliceVision

3D Computer Vision Framework

What it solves

AliceVision provides a way to reverse the process of photography by inferring the 3D geometry of a scene from a set of unordered photographs or videos. It solves the problem of creating accurate 3D reconstructions and performing camera tracking from 2D images.

How it works

It utilizes a photogrammetric computer vision framework consisting of a pipeline of state-of-the-art algorithms. These algorithms process 2D projections (photographs) to recover the lost depth information and reconstruct the 3D scene. The pipeline can be launched via Meshroom, which provides both a graphical user interface and a command-line interface for automation and customization through Python scripts.

Who it’s for

This framework is designed for researchers in academia and industry professionals who require robust, production-quality 3D reconstruction and camera tracking tools.

Highlights

  • 3D Reconstruction Pipeline: A complete framework for turning photos into 3D models.
  • Camera Tracking: Algorithms specifically for tracking camera movement within a scene.
  • Production-Ready: Developed through collaboration between academia and industry for high robustness and quality.
  • Extensible: Supports custom automation and pipeline customization via Python through Meshroom.

Related

  • Project
    alicevision/Meshroom

    A node-based visual programming framework for 3D reconstruction and computer vision, enabling users to build complex data processing pipelines using AI and photogrammetry.

  • Project
    WHU-USI3DV/VistaDream

    VistaDream is a PyTorch library that reconstructs a high‑quality 3‑D scene from a single RGB image using diffusion models and a training‑free multi‑view consistency sampler. It builds a coarse scaffold, iteratively inpaints RGB‑D views, and enforces consistency across generated views. Demos run with `python demo.py` (single view) or `python demo_sparse.py` (sparse views). Requires Ubuntu 20.04, an RTX 4090 GPU, and several pretrained models (Fooocus, Depth‑Pro, etc.). Accepted at ICCV 2025 and IEEE TPAMI 2026.

  • Project
    jiafeng5513/Evision

    A binocular vision system for camera calibration, disparity mapping, and 3D reconstruction using algorithms like ELAS and ADCensus.

  • Project
    mapillary/OpenSfM

    A Structure from Motion library in Python that reconstructs 3D scenes and camera poses from multiple images, integrating sensor data for geographical alignment.

  • Project
    colmap/colmap

    COLMAP is an open‑source Structure‑from‑Motion and Multi‑View Stereo pipeline that automatically turns collections of photos into 3‑D models. It offers a GUI, command‑line tools, Python bindings, GPU‑accelerated feature extraction, and cross‑platform binaries, making it a go‑to foundation for research and applications in computer vision, robotics, AR/VR, and any workflow that needs camera poses and dense scene geometry.