How do I quantitize 3D extrinsic calibration error?
The 3D sensor is mounted on the robot flange; the calibration plate is fixed and the robot moved around this calibration pattern to capture different views of the pattern.
The optimization algorithm was provided a rough guess extrinsic calibration matrix to make sure it converged to the right solution.
The calibration matrix found is pretty close to the guess matrix, which indicates the extrinsic calibration (Hand-Eye calibration) probably went good.
What would be the best way to estimate my extrinsic calibration matrix error? I'm looking for ideas, protocols etc.
add a comment
