6D Robotic Assembly Based on RGB-only Object Pose Estimation

TL;DR

This paper presents a novel RGB-only vision-based robotic system capable of precise 6D object pose estimation and assembly, utilizing synthetic training data and collision-free manipulation for accurate construction tasks.

Contribution

It introduces an integrated system combining monocular 6D pose estimation trained on synthetic data with a collision-free assembly method and a new calibration technique for improved accuracy.

Findings

High assembly accuracy demonstrated through experiments

Effective 6D pose estimation using synthetic images

Robustness enhanced by novel calibration method

Abstract

Vision-based robotic assembly is a crucial yet challenging task as the interaction with multiple objects requires high levels of precision. In this paper, we propose an integrated 6D robotic system to perceive, grasp, manipulate and assemble blocks with tight tolerances. Aiming to provide an off-the-shelf RGB-only solution, our system is built upon a monocular 6D object pose estimation network trained solely with synthetic images leveraging physically-based rendering. Subsequently, pose-guided 6D transformation along with collision-free assembly is proposed to construct any designed structure with arbitrary initial poses. Our novel 3-axis calibration operation further enhances the precision and robustness by disentangling 6D pose estimation and robotic assembly. Both quantitative and qualitative results demonstrate the effectiveness of our proposed 6D robotic assembly system.