基于眼固定的机器人视觉定位

摘要

针对眼固定型机器人的视觉定位问题，提出了一种基于kalman滤波的机器人无标定视觉定位方法。在摄像机和机器人坐标系不标定的情况下，采用kalman滤波算法对眼固定的图像雅克比矩阵进行在线辨识，并建立基于图像特征的最优反馈控制。利用机器人Matlab仿真工具箱建立了基于眼固定的机器人无标定视觉定位Simulink 模型，实现了6自由度机器人的视觉定位，实验结果表明了该方法的可行性和有效性。%An un-calibrated robot vision positioning method based on kalman filter is proposed focus on the robot with fixed eye. Kalman filter algorithm is adopted to on-line identify the image Jacobian matrix in the un-calibrated hand-eye coordination systems, and build the optimal feedback control based on the image features. The simulink model for robot uncalibrated vision positioning system using eye-fixed configuration is built with the robotics toolbox for Matlab. Finally, Experimental results testify the feasibility and validity of this method.