Skip to content

Latest commit

 

History

History
126 lines (91 loc) · 4.38 KB

README.md

File metadata and controls

126 lines (91 loc) · 4.38 KB

Evaluation of map-based LiDAR Localization Frameworks

Everything is hosted in Docker Containers 🐳📦

"Diese Dokumentation ist schäbig"   [ich, 2024]

Usage

Build: ./docker/build_docker.sh --dockerfile <name>

Run: ./docker/run_docker.sh --dockerfile <name>

Example

./docker/build_docker.sh --dockerfile ros1_bridge
./docker/run_docker.sh --dockerfile ros1_bridge

Tools

kitti2bag

Convert sequences of the KITTI dataset to a rosbag.
Example:

kitti2bag -t 2011_10_03 -r 0027 raw_synced .

Your file structure has to look like:

2011_10_03/  
|-- calib_cam_to_cam.txt  
|-- calib_imu_to_velo.txt  
|-- calib_velo_to_cam.txt  
|-- 2011_10_03_drive_0027_sync/

ros1_bridge

Bridge topics between ROS1 and ROS2.

  1. Start a terminal with roscore
  2. Start a second terminal
    a. source ROS1
    b. source ROS2
    c. export ROS_MASTER_URI=http://localhost:11311
    d. ros2 run ros1_bridge dynamic_bridge

bag2pcd

Convert livox-msgs to pointcloud-msgs and export pointcloud-msgs to pcd files.

Conversion:

  1. New Terminal: roslaunch livox_repub livox_repub.launch
  2. New Terminal: rosbag play /path/to/bag
  3. New Terminal: rosbag record -a

Extraction:

  1. rosrun pcl_ros bag_to_pcd <input_file.bag> <topic> <output_directory>

rosbags-convert

NO DOCKER

Convert a ROS1 bag to a ROS2 bag.

  1. Create a virtual environment (tested with python=3.10)
  2. pip3 install 'rosbags>=0.9.11'
  3. rosbags-convert --src /path/to/bag.bag --dst /path/to/output/folder/name

Localization

fast_lio_localization

  1. New Terminal: roslaunch fast_lio_localization localization_velodyne.launch map:=/datasets/map.pcd
  2. New Terminal: rosbag play /datasets/kitti_2011_10_03_drive_0027_synced.bag
  3. (optional) You can also specify an initial guess, if you are not starting at the map origin.
    New Terminal: rosrun fast_lio_localization publish_initial_pose.py x x z yaw pitch roll

liloc

Switch mapping or map-based localization with mode "lio" or "relo" in the config.

  1. Resolve an issue: - src/imageProjection.cpp → remove lines 417-420 (if is_dense statement)
  2. New Terminal: roscore
  3. New Terminal: roslaunch liloc run_lio_sam_default.launch
  4. New Terminal: rviz with specific config (launch/include/config/rviz.rviz)
  5. New Terminal: rosbag play /datasets/kitti_2011_10_03_drive_0027_synced.bag

lidar_localization_ros2

  1. New Terminal: ros2 launch lidar_localization_ros2 lidar_localization.launch.py
  2. New Terminal: rviz2 -d src/lidar_localization_ros2/rviz/localization.rviz
  3. New Terminal: ros2 bag play /datasets/KITTI/kitti_2011_10_03_drive_0027_synced/ --topics /kitti/velo/pointcloud --remap /kitti/velo/pointcloud:=/velodyne_points

DLL

Needs a map in .bt (binary octomap format), which does not have negative entries.

  1. New Terminal: ros2 launch dll kitti_launch.xml
  2. New Terminal: ros2 bag play /datasets/kitti/kitti_2011_10_03_drive_0027_synced/

hdl_localization

  1. New Terminal: roslaunch hdl_localization hdl_localization.launch
  2. New Terminal: rviz (config: /dev_ws/src/hdl_localization/rviz/)
  3. New Terminal: rosbag play /datasets/kitti/kitti_2011_10_03_drive_0027_synced.bag

BM-Loc

Downlaod the example data from the GitHub and switch the data in Bms_for_test/M2DGR_Bms/M2DGR/street_01/ with your data.
If you only have a single map, rename your map to 000.pcd and remove everything besides the 000.pcd and the CentroidCloud.pcd.

  1. New Terminal: roslaunch block_localization run_m2dgr.launch
  2. New Terminal: rosbag play /datasets/kitti/kitti_2011_10_03_drive_0027_synced.bag

range-mcl

not yet tested

LiDAR-Localization-100FPS

tbd

G3Reg

tbd

PALoc

tbd

Troubleshooting

  1. If you have problems with the attachment to a display, execute xhost + before starting the docker.

Contribute

Feel free to open an issue and a PR.
If you find a problem, feel free to solve it :)