【摘要】 口腔手术机器人手眼标定精度直接影响口腔手术机器人的工作精度。针对固定视点的口腔手术机器人，提出一种基于SVD分解的手眼标定算法，该算法无需采集姿态信息。通过纯平移运动，采集若干对坐标点计算旋转矩阵，而后通过纯旋转运动，采集若干对坐标点计算平移向量，从而完成手眼标定。在双目视觉位置定位精度为0.12 mm、机器人位置定位精度为0.1 mm、两者姿态定位精度均不定的条件下，实物实验结果表明：该算法标定平均误差的旋转部分在4 mrad以下，平移部分在0.411 mm以下，可满足口腔手术机器人手眼标定的精度要求。更多还原

【Abstract】 Oral surgery robots hand-eye calibration accuracy directly affects its working accuracy.Aiming at the eye-to-hand oral surgery robot, a hand-eye calibration algorithm based on SVD decomposition was proposed, which did not require posture information.Through pure translation motion, several pairs of coordinate points were collected to calculate the rotation matrix.Then through pure rotation motion, several pairs of coordinate points were collected to calculate the translation vector, so as to complete the hand-eye calibration.Under the condition that the binocular visual positioning accuracy is 0.12 mm, the robot positioning accuracy is 0.1 mm, and the attitude accuracy of the two is uncertain, the physical experimental results show that the rotation part of the average error of the algorithm calibration is below 4 mrad, and the translation part is below 0.411 mm, which can meet the accuracy requirements of the hand-eye calibration of oral surgery robots.更多还原