Robot manipulation grasping of recognized objects for assistive technology support using stereo vision.

Abstract

This paper demonstrates a “Grasp X ” robot arm manipulation using a 2.5D disparity map and a Scale Invariant Feature Transform (SIFT) algorithm for recognizing the target object. The SIFT algorithm is applied after stereo segmen-tation, which reduces the matching time. Due to the high inaccuracy of the stereo disparity map and the maneuverability limitation of a two-finger gripper of the robot, RT 100, a cube model is used in order to calculate the object centroid as a grasp point. 1