Implementation of an event-based algorithm for trajectory reconstruction during EDL

The exploration of celestial bodies such as the Moon, Mars, or asteroids requires reliable and fully autonomous systems to handle the critical Entry, Descent, and Landing (EDL) phase. Due to communication delays, direct ground-based control during this stage is often infeasible, making autonomous navigation solutions a key enabler for future space missions. Among emerging technologies, event-based vision stands out for its substantial advantages over traditional frame-based cameras in terms of low latency, high dynamic range, and reduced data rates. This thesis investigates the implementation of event-based vision for spacecraft navigation during lunar landing, with an emphasis on techniques for trajectory reconstruction. Following an analysis of current EDL technologies and recent mission strategies, the work introduces the principle of event-based vision and motivates its applicability to landing scenarios. A system architecture is designed to implement pose estimation and reconstruction using event-generated data. The methodology is validated using two distinct scenarios: a synthetic dataset produced with Blender to simulate descent trajectories over the lunar surface, and an analogue testbed dataset obtained from a laboratory experimental setup. These event streams are processed to extract landmarks and verify the matching methodology. The performance of the system is evaluated and compared across these two specific datasets, analyzing how different parameter settings influence both accuracy and robustness. The results demonstrate the potential of event based vision as a navigation aid in EDL, while identifying existing limitations and areas for future improvements. Ultimately, the work aims to provide a reproducible framework for assessing the effectiveness of event cameras in autonomous planetary landing scenarios, merging innovative sensing technology with practical systems-level validation.