Satellite Pose Estimation with Deep Landmark Regression and Nonlinear\n Pose Refinement

2019/08/30 by Bo Chen, Jiewei Cao, Chen, Bo +6 · 1 voice · 8 citations
Computer Science · Engineering · Physics and Astronomy · #3D pose estimation #A priori and a posteriori #Articulated body pose estimation #Artificial intelligence #Astro and Planetary Science #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Computer vision #FOS: Computer and information sciences #Landmark #Planetary Science and Exploration #Pose #Space Satellite Systems and Control #cs.CV

paper · pdf · doi:10.48550/arxiv.1908.11542

published in arXiv (Cornell University) (Cornell University) · Accepted by ICCVW 2019

arxiv created 2019/08/30 · openalex publication_date 2019/08/30 · arxiv updated 2019/09/02 · openalex created_date 2021/05/24 · openalex updated_date 2026/08/06

Abstract

We propose an approach to estimate the 6DOF pose of a satellite, relative to\na canonical pose, from a single image. Such a problem is crucial in many space\nproximity operations, such as docking, debris removal, and inter-spacecraft\ncommunications. Our approach combines machine learning and geometric\noptimisation, by predicting the coordinates of a set of landmarks in the input\nimage, associating the landmarks to their corresponding 3D points on an a\npriori reconstructed 3D model, then solving for the object pose using\nnon-linear optimisation. Our approach is not only novel for this specific pose\nestimation task, which helps to further open up a relatively new domain for\nmachine learning and computer vision, but it also demonstrates superior\naccuracy and won the first place in the recent Kelvins Pose Estimation\nChallenge organised by the European Space Agency (ESA).\n

Cited by

• FastPose-ViT: A Vision Transformer for Real-Time Spacecraft Pose Estimation
• NeRF-based Visualization of 3D Cues Supporting Data-Driven Spacecraft Pose Estimation
• Adaptive Relative Pose Estimation Framework with Dual Noise Tuning for Safe Approaching Maneuvers
• GKNet: Graph-based Keypoints Network for Monocular Pose Estimation of Non-cooperative Spacecraft
• Event-RGB Fusion for Spacecraft Pose Estimation Under Harsh Lighting
• Enabling Robust, Real-Time Verification of Vision-Based Navigation through View Synthesis
• Deterministic Object Pose Confidence Region Estimation
• A large-scale, physically-based synthetic dataset for satellite pose estimation

Discussions

• Satellite Pose Estimation with Deep Landmark Regression and Nonlinear Refinement [hn, 2 points, 0 comments]

Related

• Deep Learning for Spacecraft Pose Estimation from Photorealistic Rendering
• Bridging the Domain Gap in Satellite Pose Estimation: a Self-Training Approach based on Geometrical Constraints
• Spacecraft Pose Estimation Based on Unsupervised Domain Adaptation and on a 3D-Guided Loss Combination
• A large-scale, physically-based synthetic dataset for satellite pose estimation
• Domain Generalization for In-Orbit 6D Pose Estimation