Estimating a garment grasping point for robot

Manipulating clothing items using robots has been an active research topic for some time now. It has a lot of applications and faces many unsolved challenges. In this paper, we consider the problem of detecting a grasping point for a single-arm robot to pick up a pile of clothes for loading a washing machine. The primary goal is to have the grasped item as centralized as possible. We present a pipeline for estimating a grasping point for the robot end effector using depth camera. We analyze the depth and color information to find the optimum point based on depth and color analysis of the surface. The pipeline was tested using Doosan A0509 robot and Kinect depth camera.