Odometries

Odometries is a pure Rust library for odometry estimation and mapping.

Warning

This library is still under deep development, and the API may have breaking change at any time. If you also want to contribute to this project, welcome to open an issue or PR :)

Feautres

• Zero ROS dependency, so it is easy to use and deploy on different platforms.
• Most Generic interface for different odometry algorithms.
• Keep it Simple: This library only provides the core algorithms, you might need to manage the data flows like pointcloud stream, and use odometries to process them to get pose state.

Current status

• common wrappers and utils like Framed and Uncertained.
• The most generic ESKF framework, including State and Measurement types.
• Generic Voxelmap implementation with plane and uncertain intergration, based on slab.
• LIO: tightly‑coupled lidar imu odometry with Voxelmap map storage.
• Some examples to test the odometry algorithms.
• [-] Leg-Kilo: almost done, which benefit from LIO module. But kinematic observation works is needed.
• Fast-LIO2: need KD-Tree implementation.
• Fast-LIVO2: might reuse LIO module for lidar and imu observation, but vision implementation is needed. The crate kornia or the crate image with imageproc could help.
• Write a ROS example package using ros2-client to show how to use this library.