The exploration of Small Solar System Bodies represents one of the most relevant themes of the last two decades in the space sector. The abundance and diversity of these objects is driving a continuous interest from space actors, motivated by their scientific significance, by technological demonstrations, and by planetary defence efforts. However, small body missions still rely heavily on ground-based operations. This approach is unsuitable for their increasing number and complexity. Innovative mission concepts are characterized by more stringent scientific and technological objectives, which are challenging the current ground-based control strategies. This thesis investigates vision-based navigation methods aimed at increasing the autonomy level of spacecraft operating in close proximity of asteroids. Two feature-based approaches are considered: Visual Odometry and Simultaneous Localization and Mapping. The former focuses on estimating the spacecraft motion from an image sequence, while the latter expands the problem by including the determination of the target shape in the estimation process. Both methods rely on the output of a feature tracking algorithm that performs efficient processing of the image sequence. In addition, algorithms to autonomously estimate the rotational state of the target are implemented. In-depth analyses are carried out to evaluate the performance of the proposed methods in a variety of conditions.
L' esplorazione dei corpi minori del Sistema Solare rappresenta uno dei temi principali delle ultime due decadi nel settore spaziale. La loro abbondanza e diversità sta generando un continuo interesse da parte degli attori del settore, motivato dalla loro importanza scientifica, da dimostrazioni tecnologiche e dagli sforzi di difesa planetaria. Tuttavia, le missioni a corpi minori dipendono ancora fortemente dalle operazioni a terra. Questo approccio non è adatto al loro crescente numero e all'aumento della loro complessità. I nuovi concetti di missione sono caratterizzati da obiettivi scientifici e tecnologici sempre più stringenti, che stanno mettendo alla prova le attuali strategie basate sul controllo da terra. Questa tesi investiga metodi di navigazione ottica finalizzati ad aumentare il livello di autonomia dei satelliti che operano in prossimità di asteroidi. Due approcci basati su feature ottiche sono considerati, noti come Visual Odometry e Simultaneous Localization and Mapping. Il primo si concentra sulla stima del moto del satellite da una sequenza di immagini, mentre il secondo espande il problema includendo anche la stima della forma del corpo minore. Entrambi i metodi sono basati sull'output di un algoritmo di feature tracking che processa in modo efficiente la sequenza di immagini. In aggiunta, due algoritmi per la stima autonoma dello stato rotazionale di un corpo minore sono implementati. Analisi dettagliate sono eseguite per valutare l'accuratezza dei metodi proposti in diverse condizioni.
Autonomous vision-based navigation at small bodies with feature tracking
Piccolo, Felice
2025/2026
Abstract
The exploration of Small Solar System Bodies represents one of the most relevant themes of the last two decades in the space sector. The abundance and diversity of these objects is driving a continuous interest from space actors, motivated by their scientific significance, by technological demonstrations, and by planetary defence efforts. However, small body missions still rely heavily on ground-based operations. This approach is unsuitable for their increasing number and complexity. Innovative mission concepts are characterized by more stringent scientific and technological objectives, which are challenging the current ground-based control strategies. This thesis investigates vision-based navigation methods aimed at increasing the autonomy level of spacecraft operating in close proximity of asteroids. Two feature-based approaches are considered: Visual Odometry and Simultaneous Localization and Mapping. The former focuses on estimating the spacecraft motion from an image sequence, while the latter expands the problem by including the determination of the target shape in the estimation process. Both methods rely on the output of a feature tracking algorithm that performs efficient processing of the image sequence. In addition, algorithms to autonomously estimate the rotational state of the target are implemented. In-depth analyses are carried out to evaluate the performance of the proposed methods in a variety of conditions.| File | Dimensione | Formato | |
|---|---|---|---|
|
PiccoloFelice_PhD_Thesis.pdf
accessibile in internet per tutti
Dimensione
59.43 MB
Formato
Adobe PDF
|
59.43 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/245117