Home

Awesome

KUKA LBR iiwa Camera Calibration Tool

This package is a ROS extension for the Camera-to-Arm tool developed by Zachary Taylor. It tries to reduce the headache of manually recording the end-effector to robot transformation at each calibration pose and command the ROS node to capture the current image and save the current end-effector pose.

<!--more-->

This package is mainly implemented for KUKA LBR iiwa R820. But theoretically it can be modified with ease to fit any manipulator. Just change the image topic name and find a way to transmit a pose at each calibration view.

For video explanation visit Youtube video.

To run on KUKA LBR R820:

Sunrise side:

Add the CameraCalib.jave file to the robot applications in Sunrise OS. Complie and synchronize with the KUKA Cabinet. This file is in charge of sending the calibration poses of the end-effector to the ROS package. Follow the instructions on the SamrtPAD, enable the CCP user defined key and move the robot to a suitable place in front of the mounted camera. To save a configuration just press the user defined CCP. Don't press CCP again untill runing the ROS package.

roscore side:

  1. Download and install ROS Disrtibution.

  2. Download the kuka_camera_calib package in the ros workspace ~/ROS_ws/src/

    git clone https://github.com/elhussieny/    kuka_camera_calib

  3. catkin_make the workspace:

    catkin_make

  4. Creat an empty directory Calibration/Images inside the kuka_camera_calib directory:

    mkdir -p Calibration/Images

  5. Run the image streaming node; for example kinect2 bridge that will publish image topics:

    roslaunch kinect2_bridge kinect2_bridge

  6. Run the kuka_camera_calib node:

    rosrun kuka_camera_calib kuka_calib

  7. The image should be viewed in a window. Pressing the CCP on the KUKA SmartPAD will save the current pose of the end-effector and also save the current displayed image. The final file that contains all the poses will be saved in calib_xxx.csv file in the kuka_camera_calib/Calibration file.

MATLAB side:

  1. Open MATLAB 8.0+ and jump to the kuka_camera_calib/matlabCode/ and run the following command in the matlab command window:

    calibrateKUKA('filename.csv',chess_square_ width_in_mm)

    for instance, invoke it in this way:

    calibrateKUKA('calib_1461440868.500580.csv',27.5)

  2. Wait till the code finished and check these outputs:

    TBase- The camera to arm base transformation.

    TEnd- The end effector to checkerboard transformation.

    cameraParams- The intrinsic parameters of the camera used.

    TBaseStd- An estimation of the standard deviation of the error in the values in TBase.

    TBaseStd- An estimation of the standard. deviation of the error in the values in TEnd.

    pixelErr- Mean error of projected inliers in pixels.