Robot-camera and Hand-target Calibration in Eye-to-hand Navigation System

Abstract

In recent years, robot-assisted surgery systems have developed rapidly, and are widely used in clinics. Ensuring the accuracy of the robot-assisted surgery system is the prerequisite for the safety and stability of the system. Hand-eye calibration is necessary for an eye-to-hand surgical navigation system. However, errors exist in the data of hand-eye calibration, including the positioning error of the robot and the measurement error of the navigation system, which seriously affects the accuracy of the results of hand-eye calibration. In order to unite the coordinate systems of each device to a common coordinate system in the eye-to-hand surgical navigation system, obtain the homogeneous transformation matrix (HTM) from the positioning target to the surgical tool (hand-target), and the HTM from the camera to the base of surgical robot (robot-camera), a novel calibration method is proposed in this paper. The calibration of the HTM of hand-target is divided into the translation part and the rotation part to calibrate respectively: the multi-point method is used to calculate the translation part, and the calibration of the rotation part is summarized as an AX=XB problem. The method solves the rotation part and translation part independently, without the problem of error propagation. And the Least squares method is used to solve the HTM of robot-camera. Experimental analyses have been carried out to verify the accuracy and effectiveness of the proposed method.