Hand-eye Calibration

,
8

Have a look at the calibrateHandEye() documentation:

image

cTt is the transformation from the chessboard pattern to the camera frame.You can use findChessboardCorners() and solvePnP() to estimate the pose of the static chessboard with respect to the camera frame.

bTg is the transformation from the gripper frame to the robot base frame. The robot SDK should allow you to have this transformation.

The transformations:

  • bTt is the transformation from the calibration target to the robot base frame. It is a static transformation.
  • gTc is the transformation from the camera frame to the gripper frame. It is a static transformation, and is the transformation you want to estimate.

About your figure, you have to display the frames with respect to the same reference frame. It should look like this (from):

image
image800×600 59 KB