Contains calibration documents, logic, inputs, and outputs for several cameras that can be used with ROS.

• Python: 3.6 with pyenv

  pyenv install 3.6
  pyenv global 3.6
  

• OpenCV for Python

  pip install --upgrade pip
  pip install opencv-contrib-python
  

The pattern used is a 9x6 OpenCV chess board. If it's printed on 8.5 x 11 sheet of paper with no scaling, real-world dimensions will be 217 mm x 153 mm.

• Connect your camera and ensure it works with ROS by using Rviz
• Create a folder for your camera in this repository
• Change into the folder
• Capture 20-40 images with image_view

  rosrun image_view image_view image:=/camera/rgb/image_raw
  

  On the image view window, simply right-click and the current image will be saved to the current directory.

Run calibration.py script at the root of this repository with working directory containing the images to analyze.

• Intel Realsense D435 parameters in /intel-realsense-d435
• Primesense Carmine 1.09 parameters in /primesense-carmine-1.09

Once the camera is calibrated, markers can be detected in 3D space.

Build ar_track_alvar from source.

rosrun ar_track_alvar individualMarkers \
  _marker_size:=9.577 \
  _output_frame:="camera_rgb_optical_frame" \
  _cam_image_topic:="/camera/depth_registered/points" \
  _cam_info_topic:="/camera/rgb/camera_info" \
  _max_new_marker_error:=0.5 \
  _max_track_error:=0.5