MeshInspector/MeshLib

3D geometry processing SDK: fast mesh boolean, mesh repair, decimation, remeshing, offset, point cloud triangulation and ICP registration. C++ with Python, C#, C and JavaScript bindings.

What it solves

MeshLib is a high-performance SDK for processing, repairing, and optimizing 3D mesh and point cloud data. It addresses the common challenges of dealing with "messy" 3D scans—such as holes, self-intersections, and non-manifold geometry—to create watertight, printable, or analysis-ready models.

How it works

Built with a native C++ core, the library provides official APIs for Python, C#, C, and JavaScript/WebAssembly. It utilizes a half-edge data structure to prevent non-manifold errors by construction and offers both exact and voxel-based processing modes. For heavy workloads, it includes a CUDA module to accelerate distance maps, signed distance fields (SDF), and offsets on NVIDIA GPUs.

Who it’s for

It is designed for software engineers working in 3D scanning, metrology, medical imaging (e.g., surgical planning, orthodontics), generative design, and 3D printing/CAM software development.

Highlights

  • Comprehensive 3D Toolset: Supports mesh booleans, decimation, triangulation of point clouds, and ICP registration.
  • Robust Repair: Voxel-based repair rebuilds surfaces via signed distance fields to eliminate holes and self-intersections in a single pass.
  • Cross-Platform & Multi-Language: Runs on Windows, macOS, Linux, and WebAssembly with a single engine across five languages.
  • High Performance: Multithreaded core algorithms with GPU acceleration for specific heavy-duty operations.
  • Broad File Support: Handles various formats including STL, PLY, DICOM, and G-code.

Related

  • Project
    jasmcaus/caer

    Caer is a Python‑only, GPU‑accelerated computer‑vision library that offers fast image/video utilities (resize, colour conversion, augmentations, etc.) with a clean, type‑checked API. It can replace OpenCV for research prototypes, is installable via `pip`, and is MIT‑licensed.

  • Project
    LAION-AI/CLIP_benchmark

    A standardized evaluation framework for CLIP-like models to measure performance on zero-shot classification, retrieval, captioning, and linear probing across diverse datasets.

  • Project
    UB-Mannheim/zotero-ocr

    A Zotero plugin that uses Tesseract OCR to add a searchable text layer to scanned PDFs within a Zotero library.

  • Project
    apple-aiml-research/ml-mobileone

    MobileOne provides PyTorch implementations of five ultra‑fast CNN backbones (S0‑S4) that achieve 71‑79 % ImageNet Top‑1 accuracy with ≤ 2 ms latency on an iPhone 12 Pro. The repo includes training‑ready and inference‑ready checkpoints, CoreML models for on‑device deployment, and an iOS benchmark app (ModelBench). Re‑parameterization fuses training‑time branches into a single fast network for inference.