Collection of packages, nodes and algorithms to perform perception tasks using LiDAR and cameras.

1. Install dependencies with rosdep

rosdep install --from-path src --ignore-src -y -r

2. Build the package using colcon

If using Oak D Pro camera, install udev rules with the script ./turtlebot3_perception/debian/udev/install_udev

sudo ./install_udev

1. Set the environment variable CAMERA_MODEL to "realsense" or "oakd"
2. Launch the camera driver with

ros2 launch turtlebot3_perception camera.launch.py

1. Launch the AprilTag detection using

ros2 launch turtlebot3_perception apriltag.launch.py

• detection2landmark: a node to transform AprilTag detections in landmarks with range and bearing. See Start AprilTag landmark detection
• laserscan2lines: a node to extract lines from from LaserScan messages. See Lines from LaserScan
• landmark_simulator: a node that simulates the detection of landmarks from a list of their coordinates. See Landmarks Simulator

Run the example node with

ros2 run turtlebot3_perception laserscan2lines

You can use the function directly in your nodes putting this import in your file

from turtlebot3_perception.laserscan2lines import laserscan2lines

Launch the required nodes with

ros2 launch turtlebot3_perception apriltag.launch.py

This node simulate the detection of landmarks from a list of their coordinates. It provides range and bearing for each detected landmark and it also publishes TF to visualize in Rviz.

Example run

ros2 run turtlebot3_perception landmark_simulator --ros-args -p use_sim_time:=true

Subscriber: "ground_truth", nav_msgs/msg/Odometry

Publisher: "landmarks", landmark_msgs/msg/LandmarkArray

Static parameters:

• landmarks_file: absolute path to landmarks file. An example file is provided in landmarks.yaml
• frequency_hz: the frequency at which landmarks are published

Dynamically configurable parameters:

• field_of_view_deg
• max_range, min_range
• range_stddev
• bearing_stddev_deg