curbatpro

curbatpro is a cross-platform software library for curvature analysis in computational geometry and computer graphics. It provides algorithms for estimating and querying surface curvature, including mean curvature, Gaussian curvature, and principal curvatures, on triangulated meshes and point clouds. The library emphasizes numerical robustness and interoperability with common geometry processing pipelines and data formats.

Origins and development: curbatpro originated from an academic collaboration between geometry processing researchers and software engineers. The initial prototype focused on reliable curvature estimation from noisy data, and since its public release the project has evolved into a widely used tool with ongoing maintenance and community contributions.

Features: The library supports per-vertex and per-face curvature, discrete differential geometry operators, multiple estimation schemes, smoothing and noise-handling options, and import/export in common formats. It offers language bindings for C++ and Python, a modular API, and hooks for visualization and integration with existing workflows.

Implementation and architecture: Core components are implemented in C++ with Python bindings via pybind11. The design includes a core math layer, adapters for meshes and point clouds, and extensible IO. It emphasizes numerical stability, unit tests, and comprehensive documentation.

Applications and usage: curbatpro is used in computer-aided design, reverse engineering, medical imaging, and 3D printing to analyze curvature distributions on surfaces. Researchers use it to quantify surface features in papers, while practitioners integrate it into pipelines to assess part quality and design intent.

Reception and status: The project maintains an active open-source repository with regular releases and community involvement. Some criticisms relate to performance on very large meshes and the learning curve for new users, which contributors address through optimization work and expanded tutorials.