登录

rko_lio

PRBonnMIT活跃维护

一个不依赖传感器专门建模的鲁棒 LiDAR-惯性里程计，直接融合激光雷达与 IMU 数据估计位姿并建图，适合机器人 SLAM 场景。

A Robust Approach for LiDAR-Inertial Odometry Without Sensor-Specific Modelling

v0.4.0发布于 2026-09-01 16:24 UTC687star80fork同步于 2026-11-10 17:57 UTC
激光惯性里程计SLAM位姿估计建图机器人
对比 GitHub 官网imuinertiallidarlidar-inertial-odometrymapping

怎么上手

由 README 提炼，执行前请确认
原文 
pip install "rko_lio[all]"
rko_lio -v /path/to/data
硬件要求按 README 说明

README 未说明显存、CPU 或 GPU 型号要求

下载：源码 zip·GitHub Releases（安装包）本站不镜像安装包

项目说明

由 README 自动翻译提炼
没人工点评？写下你的经验

无需针对特定传感器建模的鲁棒 LiDAR-惯性里程计，支持 ROS1/ROS2 rosbag 直接跑

一个不依赖传感器专门建模的鲁棒 LiDAR-惯性里程计，直接融合激光雷达与 IMU 数据估计位姿并建图，适合机器人 SLAM 场景。

核心能力

  • 从 rosbag 自动读取 IMU 与 LiDAR 外参
  • 提供 pip 安装的 Python CLI 与 ROS 包两种用法
  • 支持 ROS1 和 ROS2 bag 数据加载
  • 用 rerun 可视化里程计结果
  • 支持 Humble 到 Rolling 多个 ROS 发行版

容易踩的坑

数据中缺少 IMU 与 LiDAR 外参时必须用 --config 传入含 extrinsic_imu2base_quat_xyzw_xyz 和 extrinsic_lidar2base_quat_xyzw_xyz 的配置文件；用 -v 可视化需额外安装 rerun-sdk>=0.31；ROS 话题与坐标系自动检测仍是实验性功能，失败时需手动指定 imu_topic、lidar_topic 和 base_frame

内容由模型根据 README 生成，可能不完整或不准确，以官方文档为准。看原文 README

最近发版

5 个版本v0.4.0
全部 release
  1. v0.4.0最新正式版2026-09-01 · 2026-09-01 16:24 UTC

    What's Changed

    Well. A lot.

    The most important is that there have been significant performance improvements made in this release, especially focusing on single-threaded performance. In fact, the defaults for max_num_threads has been changed to 1 from the earlier 0 (use all available) on the ROS side of things.

    More details of the improvements are covered in PRs #157 to #172. I'll just repeat the image i had in #170, which should give a quick idea of the kind of improvements we're talking about, for single threaded:

    The next main thing is better online ros node behavior.

    Failure recovery, or more precisely failure resetting, behaviors for the online ROS nodes. Earlier a registration error meant a kill. Now, odometry resets to a new identity frame and proceeds anew. Additionally, frame timing errors of 1s or more no longer mean an unrecoverable error. Instead registration is still attempted using the imu for an initial guess even on long gaps. Thanks to @Dhagash4 for this one in #174.

    And then, a quick summary of all the relevant PRs that make up this release:

    New Contributors

    Full Changelog: https://github.com/PRBonn/rko_lio/compare/v0.3.2...v0.4.0

  2. v0.3.22026-07-28 · 2026-07-28 14:20 UTC
    查看更新说明

    Minor release to pick up QoL features that just landed.

    In python, a cleanup of the rerun-based viz and a fix of the trajectory visualization (which regressed at some point with a rerun api change). The blueprint file is also no longer vendored in the package, but rather built on the fly. Cleaner.

    In ROS, a major qol feature I've been wanting for a while and harnessing the power of vibe coding, I present autodetection of lauch topics and frames. No longer do you need to specify the imu and lidar topics when you have just one of those on your system, and even the base frame is auto picked up based on the usual names. Of course you can still override flags as before to suit your needs. This is currently flagged experimental as I didn't test it extensively, but I expect there to be no issues.

    What's Changed

    Full Changelog: https://github.com/PRBonn/rko_lio/compare/v0.3.1...v0.3.2

  3. v0.3.12026-07-16 · 2026-07-16 17:53 UTC
    查看更新说明

    Minor patch to pick up changes for a ros rolling api change fix. python release will have the [all] install option not installing open3d anymore. speedup, yay.

    What's Changed

    Full Changelog: https://github.com/PRBonn/rko_lio/compare/v0.3.0...v0.3.1

  4. v0.3.02026-06-16 · 2026-06-16 15:54 UTC
    查看更新说明

    a long overdue release, packaging a bunch of performance and ux improvements. its been a pre-release for a while, and i've also been testing master's state extensively, both python and ros, so hopefully this goes smoothly.

    i think i'm slowly approaching a 1.0 release with this project (in principle, as i'll just stick with zero versioning). theres a few last things i'd want to work on here at some point, but hopefully soon

    What's Changed

  5. v0.2.02025-12-02 · 2025-12-02 08:18 UTC
    查看更新说明

    What's Changed

    New Contributors

    Full Changelog: https://github.com/PRBonn/rko_lio/compare/v0.1.7...v0.2.0

    Stay tuned for the next episode release of Dragon-- rko_lio for some more significant changes to the core library, including but not limited to performance tuning.

用户点评

登录用户的真实使用记录，可按版本和硬件筛选
共 0 条

    还没有人点评。第一个说出真实体验的人最有价值。

    README 原文

    未翻译
    在 GitHub 打开

    IEEE RA-L arXiv GitHub License GitHub last commit

    PyPI - Version ROS Package Index ROS Package Index ROS Package Index ROS Package Index ROS Package Index

    Quick Start

    Following is for the python version, see ROS for that.

    Assuming you have a rosbag (ros1/ros2) which contains a TF tree, you can run RKO-LIO through

    pip install "rko_lio[all]"
# or
pip install rko_lio rosbags "rerun-sdk>=0.31"
# data path should be a directory with *.bag files (ROS1) or a metadata.yaml (ROS2)
rko_lio -v /path/to/data

    Why pip install those three packages?

    • rko_lio -> the odometry package
    • rosbags -> required for the rosbag dataloader. Both ros1 and ros2 bags are supported!
    • rerun-sdk -> required for the optional visualizer (-v flag)

    pip install "rko_lio[all]" fetches the other optional dependencies as well.

    Check further options for the CLI through rko_lio --help.

    To dump a default config you can edit and pass with --config, run rko_lio --dump_config.

    More details are available in the Python docs.

    Extrinsics and convention

    Please note that the system needs the extrinsic to be specified between IMU and LiDAR. Either your data includes this in some format, and then the dataloaders try to automatically read it, or otherwise you can specify it in a config file (required if it's missing in the data). Pass the config file with

    rko_lio --config config_file.yaml

    This file needs two keys: extrinsic_imu2base_quat_xyzw_xyz and extrinsic_lidar2base_quat_xyzw_xyz, which must each be a list. For example: [0,0,0,1,0,0,0] for identity. Both keys are required.

    Throughout this package, I refer to transformations using transform_<from-frame>2<to-frame>. By this, I mean a transformation that converts a vector expressed in the <from-frame> coordinate system to the <to-frame> coordinate system. Mathematically, this translates to:

    $$ \mathbf{v}^{\mathrm{to}} = {}^{\mathrm{to}}\mathbf{T}_{\mathrm{from}} \mathbf{v}^{\mathrm{from}} $$

    The superscript on the vector indicates the frame in which the vector is expressed, and $^{ \mathrm{to} }\mathbf{T}_{\mathrm{from}}$ corresponds to transform_<from-frame>_to_<to-frame>.

    ROS

    Supported distros: Humble, Jazzy, Kilted, Lyrical, Rolling.

    sudo apt install ros-$ROS_DISTRO-rko-lio

    Or if you'd like to build from source, clone the repo into your colcon workspace and

    rosdep install --from-paths src --ignore-src -r -y
colcon build --packages-select rko_lio  # --symlink-install --event-handlers console_direct+

    In case you cannot system install the necessary dependencies through rosdep, you can also build the dependencies while building RKO-LIO

    colcon build --packages-select rko_lio --cmake-args -DRKO_LIO_FETCH_CONTENT_DEPS=ON

    A launch file is provided:

    ros2 launch rko_lio odometry.launch.py

    The topics and frames are autodetected (experimental). To set them yourself instead:

    ros2 launch rko_lio odometry.launch.py autodetect:=false \
    imu_topic:=<topic> lidar_topic:=<topic> base_frame:=base_link

    Those three are then the minimum you need to specify. You can specify them and other options all at once in a config file passed with config_file:=file.yaml.

    Check further launch configuration options through ros2 launch rko_lio odometry.launch.py -s

    More details are available in the ROS docs.

    The same note above about extrinsics applies here as well. Though you probably have a well defined TF tree and need not concern yourself with this (I hope).

    Citation

    If you found this work useful, please consider leaving a star :star: on this repository and citing our paper (RA-L | arXiv):

    @article{malladi2026ral,
  author      = {M.V.R. Malladi and T. Guadagnino and L. Lobefaro and C. Stachniss},
  title       = {A Robust Approach for LiDAR-Inertial Odometry Without Sensor-Specific Modeling},
  journal     = {IEEE Robotics and Automation Letters},
  year        = {2026},
  volume      = {11},
  number      = {6},
  pages       = {7420--7427},
  doi         = {10.1109/LRA.2026.3685966},
}

    The paper supplementary material page collects extra figures and explanations pertaining to the same paper.

    You can check out the branch ral_submission for the version of the code used for submission to RA-L. Please note that that branch is meant to be an as-is reproduction of the code used during submission and is not supported. The master and release versions are vastly improved, supported, and are the recommended way to use this system.

    Contributing

    My goal with this project is to have a simple LiDAR-Inertial odometry system that can work with minimal friction, in the lines of "it just works". I gladly welcome any contribution or feedback you think would help in this direction. Performance improvements or bug fixes are of course always appreciated.

    Thanks to the following contributors

    Acknowledgements

    This package is inspired by and would not be possible without the work of KISS-ICP and Kinematic-ICP. Additionally, we use and rely heavily on, either in the package itself or during development, PlotJuggler, Rerun, and of course ROS itself.

    A special mention goes out to Rerun for providing an extremely easy-to-use but highly performative visualization system. Without this, I probably would not have made a python interface at all.

    License

    This project is free software made available under the MIT license. For details, see the LICENSE file.

    相关项目

    同分类里 star 最高的项目
    对比全部 →
    rtabmap
    introlab·C++

    RTAB-Map 是一套基于外观的 SLAM 库和独立应用，支持视觉与 3D 激光建图、定位和回环检测，可在 ROS、Android、iOS 等平台运行。

    SLAM+4ROS / 机器人框架
    4093+79780.24.22026-11-09 07:23 UTC
    Prometheus
    amov-lab·C++

    面向自主无人机的开源软件栈，集成 PX4、MAVROS、Gazebo 仿真与视觉 SLAM，帮开发者快速搭建无人机定位、控制与仿真验证环境。

    无人机+4ROS / 机器人框架
    3298+4468v1.02026-04-03 05:10 UTC
    MASt3R-SLAM
    rmurai0610·Python

    基于 MASt3R 三维重建先验的实时稠密 SLAM 系统，能在单目视频流中同步完成相机位姿估计与稠密三维重建，适合机器人和 AR 场景。

    稠密SLAM+4ROS / 机器人框架
    3242+53812025-11-09 07:36 UTC
    open_vins
    rpng·C++

    开源视觉惯性导航研究平台，提供 MSCKF 与 EKF 定位、传感器标定和 VIO 算法实现，方便机器人研究者做 SLAM 与状态估计实验。

    视觉惯性导航+4ROS / 机器人框架
    3176+3950v2.72025-11-30 08:23 UTC