基于点特征的机器人视觉定位研究

Positioning of Robot Vision Based on Point Features

本文在总结过往对机器人视觉定位的研究成果上,提出一种在内参数为五参数的前提下,利用图像中的点特征对机器人进行定位,通过计算可以获得机器人的位姿信息。先是利用点特征方法获得摄像机位姿信息,再利用摄像机与机器人本体之间的固有位置关系获得机器人的位姿信息。实验验证了该方法的有效性。

This paper summarizes the results of previous studies on robot vision positioning on the premise of a proposed five parameters, including the parameters, the use of image characteristics of the robot to Focate the point, can be obtained by calculating the position and orientation information of the robot. First, the use of point feature information obtained camera pose, then use the proper positional relationship between the camera and the robot body to obtain information on the robot pose. Experiments demonstrate the effectiveness of this method.