This paper introduces a novel method using chaser spacecraft image processing and Convolutional Neural Networks (CNNs) to detect structural markers on the European Space Agency's (ESA) Environmental Satellite (ENVISAT) for safe de-orbiting. Advanced image pre-processing techniques, including noise addition and blurring, are employed to improve marker detection accuracy and robustness. Initial results show promising potential for autonomous space debris removal, supporting proactive strategies for space sustainability. The effectiveness of our approach suggests that our estimation method could significantly enhance the safety and efficiency of debris removal operations by implementing more robust and autonomous systems in actual space missions.
Markers Identification for Relative Pose Estimation of an Uncooperative Target
2024
2024 AAS/AIAA Astrodynamics Specialist Conference
Preprint
Electronic Resource
English
Mapping and Pose Estimation of an Uncooperative Rotating Target for Spacecraft Proximity Operations
Springer Verlag | 2021
|Uncooperative pose estimation with a LIDAR-based system
Online Contents | 2014
|