ouguangjun/Leg-KILO

Leg-KILO: Robust Kinematic-Inertial-Lidar Odometry for Dynamic Legged Robots

What it solves

Leg-KILO (incorporating kilo-map) provides a real-time LiDAR-based SLAM (Simultaneous Localization and Mapping) system designed to be robust against aggressive platform dynamics, such as those experienced by legged robots. It solves the problem of maintaining accurate odometry and global consistency in both structured and unstructured environments while handling high-dynamic motion distortion.

How it works

The system employs a multi-stage approach to fuse sensor data and optimize the map:

  • Two-Stage ESKF Frontend: It uses an Error-State Kalman Filter (ESKF) to fuse LiDAR and IMU data. Stage 1 performs incremental per-point updates to compensate for motion distortion. Stage 2 uses an Iterated ESKF (IESKF) over the full frame to refine global consistency.
  • Hybrid Feature Gaussian Voxel Map: It maintains planar and NDT-variant features in a voxel map to improve feature utilization across different environment types.
  • Factor Graph Backend: A Ceres-based factor graph optimization tightly couples loop-closure constraints with odometry factors to reduce global drift. It utilizes small_gicp and KISS-Matcher for loop-closure verification.
  • Cross-Platform Integration: The system is built for ROS 1 and ROS 2, supporting various LiDAR sensors (Velodyne, Ouster, Hesai, Livox).

Who it’s for

This project is for robotics developers and researchers working with legged robots or other platforms experiencing high-dynamic motion who need robust real-time 3D mapping and localization.

Highlights

  • High-Dynamic Robustness: Specifically designed to handle the aggressive movements of legged robots.
  • Hybrid Mapping: Combines planar and NDT-variant features for versatility in structured and unstructured settings.
  • Real-Time Visualization: Integrated with Iridescence for 3D debugging and visualization.
  • Broad Sensor Support: Compatible with multiple major LiDAR brands and ROS distributions.

Related

  • Project
    PRBonn/rko_lio

    A robust LiDAR-Inertial Odometry system that provides real-time position and orientation tracking without requiring sensor-specific modeling.

  • Project
    NVIDIA-ISAAC-ROS/isaac_ros_visual_slam

    A GPU-accelerated ROS 2 package for visual simultaneous localization and mapping (VSLAM) that provides high-precision odometry for robots using stereo cameras and IMUs.

  • Project
    SteveMacenski/slam_toolbox

    A 2D SLAM library for ROS 2 that enables mobile robots to build and refine maps of dynamic environments using pose-graph optimization.

  • Project
    PRBonn/kiss-slam

    A simple and robust 3D LiDAR SLAM system designed for accurate simultaneous localization and mapping with high generalization capabilities.

  • Project
    sjtuyinjie/Ultra-Fusion

    A resilient, tightly-coupled multi-sensor SLAM framework that maintains accurate localization for wheeled, legged, and aerial robots even during sensor degradation or calibration errors.