An Efficient Probabilistic 3D Mapping Framework That Embraces the Unknown.

UFO is a high-performance, modular 3D mapping framework designed for modern robotics and computer vision. It provides a robust probabilistic representation of space that explicitly handles free, occupied, and unknown regions, making it ideal for path planning and exploration in complex environments.

Key Features

Modern C++23: Leverages the latest language features for maximum performance and readability.
Probabilistic Mapping: Accurately represents uncertainty in 3D environments.
High Performance: Optimized with SIMD (SSE2, AVX, etc.) and Link Time Optimization (LTO).
Modular Architecture: Easy to extend or use individual components like Core, Math, and Utility.
Cross-Platform: Supports Linux, macOS, and Windows.

Quick Start

Prerequisites

CMake (v3.23 or newer)
C++23 Compatible Compiler (GCC 14+, Clang 18+, MSVC 19.40+)
Dependencies (automatically fetched if not found):
- Catch2 (Testing)
- CLI11 (CLI Parser)
- Doxygen (Documentation)

Build Instructions

# Clone the repository
git clone https://github.com/UnknownFreeOccupied/ufo.git
cd ufo
 
# Configure and build
cmake -B build -DUFO_DEV_MODE=ON
cmake --build build -j$(nproc)

Running the App

The main ufo utility provides system diagnostics and diagnostic information about the build environment:

./build/apps/ufo --version

Project Structure

lib/core: Fundamental types and core mapping logic (Confidence, Semantic, etc.).
lib/math: Optimized linear algebra and geometric primitives.
lib/utility: Threading, string manipulation, and system helpers.
[apps/](apps): Command-line tools and applications.
[tests/](tests): Global test suite (Catch2).

Documentation

API documentation can be generated using Doxygen:

cmake -B build -DUFO_BUILD_DOCS=ON

cmake --build build --target Docs

The output will be available in build/docs/html/index.html.

License

UFO is released under the BSD 3-Clause License.