pcl_tutorial

Tutorial for using Point Cloud Library (PCL) with ROS 2.

Code adapted for ROS 2 from ROS Industrial: Building a Perception Pipeline.

Getting Started

Install ROS 2, create a workspace then install the following packages:

sudo apt install ros-galactic-pcl-ros
sudo apt install ros-galactic-pcl-conversions

Clone this repo into a ROS 2 workspace and build:

source <ROS_WS>/install.setup.bash
cd <ROS_WS>/src
git clone https://github.com/adrian-soch/pcl_tutorial
colcon build

Running the code

After building the easiest way to start the node is through the launch file, change the paramters inside processing_node.launch.py then:

ros2 launch pcl_tutorial processing_node.launch.py

Then start a PointCloud2 Publisher, this can be from a real hardware, from a simulation, or a rosbag. To play a rosbag that has pointcloud data recorded:

ros2 bag play <PATH_TO_BAG>

Optionally, you can pause rosbags, or play them 1 message at a time:

ros2 service call /rosbag2_player/toggle_paused rosbag2_interfaces/TogglePaused

ros2 service call /rosbag2_player/play_next rosbag2_interfaces/PlayNext

Visualization

1. To view the point cloud topics, run rviz2 in a new terminal.

2. Click Add near the bottom right, select the By topic tab, and then select the point cloud 2 topic that you want to see.

3. Go to Global Options -> Fixed Frame and select the frame the data is in. Otherwise there will be a Global Status error that prevents you from seeing the data.

4. To make LiDAR points easier to see go to PointCloud2 -> Style and select Points from the dropdown menu. Feel free to try changing any other settings

5. You can save the current configuration and rviz2 will open to this configuration by default. This can be changed at any time.

Debugging

An easy way to debug ROS 2 C++ programs is through VSCode.

1. Build the package with simlinks, and the debug flag
2. Run the ROS node with gdb prefix (or add to launch file)
3. Add the following to the launch.json' in the .vscode` folder

colcon build --packages-select lesson_perception --symlink-install --cmake-args -DCMAKE_BUILD_TYPE=RelWithDebInfo

ros2 run --prefix 'gdbserver localhost:3000' lesson_perception perception_node

Create VSCode `launch.json`
{
    "version": "0.2.0",
    "configurations": [
        {
            "name": "ROS2 C++ Debugger",
            "request": "launch",
            "type": "cppdbg",
            "miDebuggerServerAddress": "localhost:3000",
            "cwd": "/",
            "program": "/home/<PATH_TO_ROS_WS>/install/<PACKAGE>/lib/<PACKAGE>/<EXECUTABLE>"
        }
    ]
}

# Run debugger in vscode

Resources

• Info on PCL, or for use witout ROS see : https://pcl.readthedocs.io/projects/tutorials/en/latest/.
• Splitting ROS nodes into multiple files (compared to the basic tutorial): https://answers.ros.org/question/380079/how-to-write-header-file-for-ros-2-publisher/.