Image-to-Action Translations Based Object Grasping Strategy without Depth Information and Robot Kinematics Analysis

The main objective of this paper is to utilize only RGB images to let a robotic arm grasp a target object without the related 3D position information. The advantages of the proposed method include image-to-action translations that apply to a class of general robotic arms, and the mathematical analys...

Full description

Saved in:
Bibliographic Details
Main Authors: Lin, Yu-Wei, Chien, Yi-Hsing, Hsu, Min-Jie, Wang, Wei-Yen, Lu, Cheng-Kai, Hsu, Chen-Chien
Format: Conference Proceeding
Language:English
Subjects:
Deep learning
Grasping
image-to-action translations
Kinematics
Mathematical analysis
object grasping strategy
proximal policy optimization (PPO)
Reinforcement learning
Robot kinematics
Three-dimensional displays
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The main objective of this paper is to utilize only RGB images to let a robotic arm grasp a target object without the related 3D position information. The advantages of the proposed method include image-to-action translations that apply to a class of general robotic arms, and the mathematical analysis of the inverse kinematics is not necessary. We employ the RGB images and the proximal policy optimization (PPO) algorithm to train the reinforcement learning network in the Gazebo simulated environment. Finally, an illustrative example shows how effective the proposed strategy is.
ISSN:2575-8284
DOI:10.1109/ICCE-Taiwan58799.2023.10226979