Skip to content

leggedrobotics/rsl_panoptic_mapping

Repository files navigation

RSL Panoptic Mapping

overview.png

RSL Panoptic Mapping is a ROS package for advanced environment understanding and navigation. It uses semantic segmentation and dynamic object tracking to create detailed panoptic maps, distinguishing between movable and static objects.

Features

  • Semantic Layer Generation: Generates semantic layers from dynamic and static point clouds.
  • Dynamic Mapping and Tracking: Separates point clouds into labeled groups for real-time tracking.
  • Traversability Cost Layer: Assigns navigation costs based on semantic information.
  • Occupancy Map: Reflects the current state of the environment with both static and dynamic objects.
  • Semantic 2D Map Integration: Enhances environmental understanding using tools like GridMap.
  • Filtering and Fusion: Applies filters to improve map accuracy.
  • ROS Integration: Fully integrated with ROS.

Please check the subpackages README.md for more information

Installation

Ensure ROS is installed. This package depends on several external ROS packages and libraries.

Dependencies

  1. GridMap
    • For semantic 2D map generation.
    git clone https://github.com/ANYbotics/grid_map.git
  2. Ground Plane Removal
    • Essential for dynamic object segmentation.
    git clone [email protected]:leggedrobotics/tree_detection.git
  3. FKIE Message Filters
    sudo apt install ros-noetic-fkie-message-filters

Usage

  1. Launch your sensors (camera and LiDAR) and publish the robot state.
  2. Launch your segmentation node or use rsl_panoptic.
  3. Launch the dynamic mapping node:
    roslaunch dynamic_mapping dynamic_mapping.launch
  4. Launch the Panoptic Gridmap:
    roslaunch panoptic_gridmap semantic_map.launch

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published