JStech / moveit_cal_simulation

18 stars 7 forks source link

Running MoveIt Calibration in simulation

Available on Docker

If you prefer, you can skip the steps below and use the Docker image instead.

Avoid OpenCV 3.2

OpenCV 3.2, which is what ships with Ubuntu 18.04, has a buggy ArUco pose detection implementation. The instructions below assume you have OpenCV 3.4 installed in /usr/local/. OpenCV 3.4 can be easily installed as follows:

git clone https://github.com/RoboticsYY/opencv_deb_install.git
sudo dpkg -i opencv_deb_install/OpenCV-3.4/OpenCV-3.4.5-x86_64-*

Downloading and building MoveIt Calibration and simulation packages

If you haven't already, install catkin:

sudo apt install ros-melodic-catkin python-catkin-tools

The simulation also requires Gazebo:

sudo apt install ros-melodic-gazebo-ros ros-melodic-gazebo-plugins ros-melodic-gazebo-ros-control \
  ros-melodic-joint-state-controller ros-melodic-position-controllers ros-melodic-joint-trajectory-controller

Set up a new workspace using the included moveit_cal_simulation.rosinstall:

mkdir -p ~/ws_moveit_cal/src
cd ~/ws_moveit_cal/src
wstool init .
wstool merge -t . https://raw.githubusercontent.com/JStech/moveit_cal_simulation/main/moveit_cal_simulation.rosinstall
wstool update -t .
rosdep install -y --from-paths . --ignore-src --rosdistro melodic

Configure and build:

cd ~/ws_moveit_cal
catkin config --extend /opt/ros/melodic --cmake-args -DOpenCV_DIR=/usr/local/share/OpenCV -DCMAKE_BUILD_TYPE=Release
catkin build
roslaunch panda_moveit_config demo_gazebo.launch

From there, follow the tutorial.