# SarahWeiii/CoACD

[SIGGRAPH2022] Approximate Convex Decomposition for 3D Meshes with Collision-Aware Concavity and Tree Search

Repository: https://github.com/SarahWeiii/CoACD
Canonical: https://ross.abutalabs.com/products/coacd
Homepage: https://colin97.github.io/CoACD/
Language: C++
License: MIT
License Family: permissive
Topics: convex-decomposition, collision, convex
Last push: 2026-08-19T04:28:56+00:00

## Health v2 (maintenance only)
Score: 98/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 98, longevity 100
- inputs: {"age_days": 1583, "days_push": 14, "days_rel": 14, "gap_med": 5, "n_releases_24m": 12}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1134, forks 152 (observed 2026-08-28T04:03:43.168871+00:00)

## What it is
CoACD is a C++ library (with Python bindings and a Unity package) that performs approximate convex decomposition of 3D triangle meshes using a collision-aware concavity metric and tree search, published at SIGGRAPH 2022. It splits meshes into nearly convex components whose convex hulls preserve collision conditions for efficient physics simulation and interaction.

## Use cases
- decompose a 3D mesh into convex parts for collision detection
- prepare game engine assets for physics simulation
- generate collision hulls for robot simulation environments
- convert scanned or CAD meshes into convex components with real-world metric thresholds
- preprocess meshes for articulated object simulation in Unity

## When to choose
- you need collision-aware convex decomposition that preserves fine details like slots or holes
- you want a pip-installable Python API or Unity package for mesh decomposition
- you need intersection-free convex hulls without voxelization errors
- you work with real-world scale meshes and want thresholds in meters

## When to avoid
- you need exact convex decomposition rather than an approximation
- your input meshes are not triangle meshes and you cannot triangulate them
- you need manifold mesh repair or low-poly conversion (use the companion PaMO library instead)

## Facets
- artifact type: library
- maturity: active
- function: graphics, simulation, machine-learning, cli
- domain: computer-vision, graphics, robotics, simulation
- platform: windows, python, cpp, cross-platform
- tags: convex-decomposition, 3d-meshes, collision-detection, physics-simulation, siggraph, unity, geometry-processing, game-development, linux, macos

## Member repositories
- SarahWeiii/CoACD (main) score 98

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:43.168871+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:36:53.328213+00:00, confidence not recorded.
  - readme: https://github.com/SarahWeiii/CoACD (fetched 2026-08-28T04:03:43.168871+00:00, sha 6d493608a078)
  - homepage: https://colin97.github.io/CoACD/ (fetched 2026-08-29T12:41:47.828656+00:00, sha 610b7716737d)
- Data as of 2026-08-30T08:39:29.467469+00:00.