A feasible vision-based measurement method for robot orientation error

A vision-based measurement method for robot orientation error is presented. The measurement can be completed by performing extrinsic parameters calibration twice, before and after the movement of the robot end-effector. It only needs a camera and a planar calib board. The process of the measurement is easy and quick with a high degree of accuracy. Experiments show that the angle error of the proposed method is lower than 0.1, and the orientation error representation is robust to the error produced in the measurement process.