Skip to content

Latest commit

 

History

History
9 lines (6 loc) · 1.61 KB

README.md

File metadata and controls

9 lines (6 loc) · 1.61 KB

arXiv

SPIN: An Open Simulator of Realistic Spacecraft Navigation Imagery

Data acquired in space operational conditions is scarce due to the costs and complexity of space operations. This poses a challenge to learning-based visual-based navigation algorithms employed in autonomous spacecraft navigation. Existing datasets, which largely depend on computer-simulated data, have partially filled this gap. However, the image generation tools they use are proprietary, which limits the evaluation of methods to unseen scenarios. Furthermore, these datasets provide limited ground-truth data, primarily focusing on the spacecraft's translation and rotation relative to the camera. To address these limitations, we present SPIN (SPacecraft Imagery for Navigation), an open-source realistic spacecraft image generation tool for relative navigation between two spacecrafts. SPIN provides a wide variety of ground-truth data and allows researchers to employ custom 3D models of satellites, define specific camera-relative poses, and adjust various settings such as camera parameters and environmental illumination conditions. For the task of spacecraft pose estimation, we compare the results of training with a SPIN-generated dataset against existing synthetic datasets. We show a 50% average error reduction in common testbed data (that simulates realistic space conditions).

Paper can be read here: https://arxiv.org/abs/2406.07500

Demo

We will soon share a functional version and a roadmap with future updates.