3D Primitive Approximation Method for Model-less Determining of Grasping Parameters

Abstract

In order to realize automated picking robot, it is an important task to determine the grasping parameters (position/direction/angle) of the object. In this paper, we propose a method for approximating an object with primitive shape to determine the grasping parameters. Our method applies “object primitive” (for example, hexahedrons, cylinders, and spheres) to the object by using a 3D-deep neural network (DNN) on the surface of the object. Then, we estimate the grasping parameters based on preset grasping rules. The success rate of approximating the object primitive with our method was 94.7%. This result is 6.7% higher than the 3D ShapeNets method using 3D-DNN. Also, as an experimental result of grasping simulation using Gazebo, the success rate of grasping with our method was 85.6%. This result is 17.8% higher than the GPD method using DNN.