摘要
Abstract
Hand-eye calibration is a key technology in the field of robot vision,which mainly includes two calibration methods,named eye in hand and eye out of hand.The purpose of calibration is to obtain the pose relationship between the camera coordinate system and the robot coordinate system,reconstruct the three-dimensional scene according to the images obtained by the camera,and calculate the conversion parameters between the two coordinate systems,the results of which will directly affect the accurate operation of the robot.Due to the limitations of the current lens technology and machine error,nonlinear distortion occurs to the image acquired by the camera,and the original hand-eye calibration method has the disadvantage of unsatisfied accuracy.In view of this,a nonlinear hand-eye calibration method for industrial robots is proposed,and the matrix parameters are adapted to the distortion in different regions of the camera lens by the convolution principle to reduce the impact of camera lens distortion.Experimental verification shows that the method can improve the accuracy of the results of hand-eye calibration,its average error of calibration is 0.4 mm,and it is more stable and can meet the needs of more scenes.关键词
手眼标定/机器人视觉/非线性畸变/卷积原理/九点标定法/旋转矩阵参数Key words
hand-eye calibration/robot vision/nonlinear distortion/convolution principle/nine-point calibration/rotation matrix parameter分类
信息技术与安全科学
