Repo for Spring 2022 RSP Course Project.
To install the package, clone the repository into a catkin workspace. This package requires that the physical turtlebot has the standard turlebot3 and raspicam packages installed on it.
For this project, we produced the code necessary to have a turtlebot use its raspicam to detect and follow a fiducial marker (AR tag), both in real life and in gazebo. To do so we have created the following packages (descriptions in Package Description section):
Camera Related
- turtlebot_camera_bringup
Hardware Related
- turtlebot_target_follow
Simulation Related
- turtlebot3_project_bringup
- turtlebot_description
- turtlebot_aruco_driver
turtlebot_camera_bringup
This package contains launch files to start the raspicam on the turtlebot and launch aruco AR tag detection (using aruco_detect package) on the main pc.
turtlebot_target_follow
This package contains the hpp and cpp code for the turtlebot target following. It is the main package of the project.
It parses the /fiducial_transforms (transformation from the raspicam on the turtlebot to detected fiducial markers) rostopic to obtain the AR tag transformation and publishes wheel velocity commands for both physical and simulated turtlebots to the apprioprate /cmd_vel rostopics based on the AR tag transformation.
The package minimizes the angle and maintans a distance of ~1 meter between the turtlebot and the detected fiducial transform. If the robot loses the fiducial marker it will rotate in place with a directionality derived from the horizontal offset it last detected the transform. If fiducial transform gets too close to the camera (< 0.8 meters from the camera) the turtlebot will reverse in order to maintain ~1 meter distance.
turtle_Project_simulation (Directory)
turtlebot_description
This package contains files required for 3D model descriptions for the turtlebots and AR tag.
turtlebot3_project_bringup
This package contains launch files required to spawn a follower turtlebot and AR tag leader Turtlebot in gazebo.
turtlebot_aruco_driver
This package contains launch file to run the aruco AR tag detection purely in simulation.
Running the tracker on the physical Turtlebot
-
Ensure the ROS_MASTER_URI and ROS_HOSTNAME are appropriately set on the turtlebot and main pc such that the ROS_MASTER_URI matches on both machines and the ROS_HOSTNAME matches the machines IP address
-
Run the following commands in your main pc's terminal:
echo 'export LDS_MODEL=LDS-01' >> ~/.bashrc
echo "export TURTLEBOT3_MODEL=burger" >> ~/.bashrc
If your model is the 'waffle' model, replace 'burger' in the line above with 'waffle'.
- In workspace, run:
The following instructions assume that a catkin workspace has been created at $WORKSPACE and the source path is present at $WORKSPACE/src
catkin build turtlebot_target_follow
source $WORKSPACE/devel/setup.bash
# Replace $WORKSPACE with the appropriate worksapce path
-
Run roscore on either turtlebot or your machine (depending on which has the MASTER_URI)
-
ssh to turtlebot and run:
roslaunch turtlebot_target_follow turtlebot_bringup.launch
This will launch the raspicamera, do the image transport needed for the ar tag processing node, and launch the turtlebot3_robot.bringup which starts all the motor drivers, gryo, etc on the turtlebot.
- On main computer run:
roslaunch turtlebot_target_follow target_follow_start.launch
This starts the aruco package which handles the ar tag transforms, gazebo client with a simulated turtlebot mimicing the real one, and the target_follow_node which reads in the ar tag transform and sends appriopriate motor commands to the cmd_vel topic for the turtlebot and mimic to move.
To change the size of the ar tag to match that of your actual ar tag size change the fiducial_length rosparam via:
roslaunch turtlebot_target_follow tartget_follow_start fiducial_length:= #tag side length in meters
To do a gazebo simulation-only version of the target tracking where the turtlebot follows a teleoperated turtlebot with an AR tag, run:
roslaunch turtlebot_target_follow follower_simulation.launch
To run ar tag vizualiation seperately
Launch camera
roslaunch raspicam_node camerav2_1280x960.launch
Image transpost converts the images into suitable compressed format to use for ar tag tracking
rosrun image_transport republish compressed in:=/raspicam_node/image raw out:=/raspicam_node/image/image_repub
Launches Ar Tag Track
roslaunch aruco_detect aruco_detect.launch transport:=compressed verbose:=true
Use rqt_gui to visualize
rosrun rqt_gui rqt_gui