This small toolset allows to integrate SLAM solution provided by GLIM with HDMapping. This repository contains ROS 2 workspace that :
- submodule to tested revision of GLIM
- a converter that listens to topics advertised from odometry node and save data in format compatible with HDMapping.
sudo apt install -y nlohmann-json3-devhttps://koide3.github.io/glim/To install the common dependencies needed for GLIM, follow the instructions provided in the official documentation:
https://koide3.github.io/glim/installation.htmlThis section provides a quick guide to run the project, including example configuration changes and launch instructions.
https://koide3.github.io/glim/quickstart.htmlClone the repo
mkdir -p /test_ws/src
cd /test_ws/src
git clone https://github.com/marcinmatecki/GLIM-to-HDMapping.git --recursive
cd ..
colcon buildPrepare recorded bag with estimated odometry:
In first terminal record bag:
ros2 bag record /glim_ros/aligned_points_corrected /glim_ros/odom_correctedand start odometry:
cd /test_ws/
source ./install/setup.sh # adjust to used shell
ros2 run glim_ros glim_rosbag <path_to_rosbag>cd /test_ws/
source ./install/setup.sh # adjust to used shell
ros2 run glim-to-hdmapping listener <recorded_bag> <output_dir>Download the dataset from NTU-VIRAL For this example, download eee_03.
rosbags-convert --src {your_downloaded_bag} --dst {desired_destination_for_the_converted_bag}ros2 bag record /glim_ros/aligned_points_corrected /glim_ros/odom_corrected {your_directory_for_the_recorded_bag}src/GLIM-to-HDMapping/src/glim/config/config_ros.jsonChanges:
Topics:
"imu_topic": "/os1_cloud_node1/imu",
"points_topic": "/os1_cloud_node1/points",
"image_topic": "/image",cd /test_ws/
source ./install/setup.sh # adjust to used shell
ros2 run glim_ros glim_rosbag {path_to_bag_file} In the terminal where the ros record is, interrupt the recording by CTRL+C
Do it also in ros launch terminal by CTRL+C.cd /test_ws/
source ./install/setup.sh # adjust to used shell
ros2 run glim-to-hdmapping listener <recorded_bag> <output_dir>