An open-source, node-based visual programming framework for creating and executing complex 3D reconstruction and computer vision pipelines.
Meshroom is an open-source, node-based visual programming framework for creating and executing complex data processing pipelines, with a primary focus on 3D reconstruction and computer vision. It allows users to visually design workflows where each node performs a specific operation, enabling the transformation of 2D image collections into detailed 3D models, camera tracking, panorama stitching, and other photogrammetric tasks. The framework efficiently manages computation through smart caching and supports both local and distributed execution on render farms.
Researchers, visual effects artists, archaeologists, surveyors, and developers working in photogrammetry, 3D scanning, computer vision, and digital heritage who need to create reproducible, complex image processing pipelines without deep programming.
Developers choose Meshroom for its intuitive visual programming interface that abstracts complex computer vision algorithms, its efficient smart caching system that minimizes recomputation, and its extensible plugin architecture that allows integration of both research and production-grade tools into a single, manageable workflow.
Node-based Visual Programming Toolbox
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
The node-based graph editor allows users to design and manipulate complex processing pipelines visually, reducing the need for coding expertise as highlighted in the Concepts section.
Attribute modifications trigger selective invalidation of downstream nodes while preserving cached results, minimizing redundant computation in iterative workflows, a core feature described in the README.
Support for local and render farm execution enables parallel processing of large image datasets, making it suitable for handling massive projects with distributed resources.
Built-in 2D and 3D viewers provide immediate visual feedback on images, depth maps, and models during pipeline execution, enhancing user interaction and debugging.
Official plugins like AliceVision for photogrammetry and community plugins for AI segmentation allow easy integration of advanced algorithms, as shown in the bundled and other plugins sections.
Mastering the node system and numerous parameters requires substantial time, especially for users new to photogrammetry and computer vision, beyond simple drag-and-drop tools.
Installing from source or adding custom plugins involves multiple steps and dependencies, as documented in INSTALL.md and INSTALL_PLUGINS.md, which can be barrier for quick adoption.
Focused on batch processing pipelines, it's not designed for applications needing instant 3D reconstruction updates or interactive editing, restricting use in live scenarios.
Some plugins, like the MicMac integration, lack full support for Meshroom's caching and invalidation features, leading to potential inefficiencies and compatibility issues.