Deep Learning for Spacecraft Pose Estimation from Photorealistic Rendering

TL;DR

This paper presents a deep learning approach for spacecraft pose estimation using a photorealistic simulator built on Unreal Engine 4, achieving competitive results in ESA challenge and demonstrating transferability to real space images.

Contribution

It introduces URSO, a novel photorealistic simulator for spacecraft images, and a deep learning framework based on orientation soft classification for pose estimation.

Findings

Achieved 3rd place on synthetic and 2nd on real ESA challenge datasets.

Demonstrated effective transfer of models trained on URSO to real space images.

Analyzed architectural and training factors impacting model performance.

Abstract

On-orbit proximity operations in space rendezvous, docking and debris removal require precise and robust 6D pose estimation under a wide range of lighting conditions and against highly textured background, i.e., the Earth. This paper investigates leveraging deep learning and photorealistic rendering for monocular pose estimation of known uncooperative spacecrafts. We first present a simulator built on Unreal Engine 4, named URSO, to generate labeled images of spacecrafts orbiting the Earth, which can be used to train and evaluate neural networks. Secondly, we propose a deep learning framework for pose estimation based on orientation soft classification, which allows modelling orientation ambiguity as a mixture of Gaussians. This framework was evaluated both on URSO datasets and the ESA pose estimation challenge. In this competition, our best model achieved 3rd place on the synthetic test set and 2nd place on the real test set. Moreover, our results show the impact of several architectural and training aspects, and we demonstrate qualitatively how models learned on URSO datasets can perform on real images from space.