Part of my thesis. My ROS2 workspace for simulations. I'm using the latest Galactic version of ROS2.
Info about the nodes in each package can be found in the README files of each package folder (Click the links in the titles).
The stereo_cam package contains scripts and simulation environments for extracting 3D pointclouds from a stereoscopic camera. Simulations are set up on Gazebo and the results presented on RViz2.
The pose_estimation package contains usefull methods for estimating the pose of the target. Starting from the easiest method with prior knowledge about the target considered, I move towards a method involving more and more autonomy.
Software tracks ArUco markers on target using a calibrated camera and extracts pose. To test the usage you can use the aruco.launch.py launch file and test the pose estimation from the computers built-in from the chaser camera in a Gazebo simulation environment.
ros2 launch pose_estimation aruco.launch.pyAutomatically loads lighting and a chessboard in a Gazebo world. The script automaticaly moves the chessboard arround, takes pictures using the chasers camera and creates the calibration parameters in a JSON file. The whole process is automated and relatively quick (see #1).
Run from the workspace directory
colcon build --symlink-installsource ./install/setup.bash
ros2 launch <package_name> <launch_file.py>
ros2 run <package_name> <node.py>