CBL - Campus del Baix Llobregat

Projecte llegit

Títol: Image Compression Strategies for Satellite Earth Observation: Constraints, Evaluation and Mission-Oriented Recommendations

Estudiants que han llegit aquest projecte:

Director/a: PARK, HYUK

Departament: FIS

Títol: Image Compression Strategies for Satellite Earth Observation: Constraints, Evaluation and Mission-Oriented Recommendations

Data inici oferta: 03-02-2025     Data finalització oferta: 03-10-2025


Estudis d'assignació del projecte:
    DG ENG AERO/TELEMÀT
Tipus: Individual
 
Lloc de realització: EETAC
 
Segon director/a extern: Luis Contreras
 
Paraules clau:
Remote sensing, Image compression, CubeSats, Bandwidth limitation
 
Descripció del contingut i pla d'activitats:
- Objectives: The goal is to develop firmware that ensures reliable command, control, and data retrieval for the On-Board Computer (OBC) of the 3Cat8 satellite. This involves testing and verifying the functionality of communication protocols. Additionally, the project will focus on implementing and validating camera capture functionality through the PoCat VGACam, as well as commanding, configuring, and retrieving telemetry from the EPS.
- Methodology: The first step will involve validating the interfaces of the On-Board Computer (OBC) using hardware tools and debugging equipment to ensure proper functionality. This process will include a review of relevant literature, such as the Endurosat OBC user manual. The project will advance into developing and testing of firmware to enable image capture through the PoCat VGACam. Furthermore, commands will be implemented to retrieve telemetry data from the EPS, verifying its integration and performance. Finally, testing will be carried out to identify and resolve any potential issues, ensuring the robustness and reliability of the developed firmware.
- Expected results: The expected results include a fully functional and validated firmware capable of executing reliable command, control, and data retrieval operations for the 3Cat8 OBC. Additionally, comprehensive documentation of the development and validation process will be provided, serving as a valuable reference for future OBC stages.
 
Overview (resum en anglès):
Satellite imaging has been a central capability of space systems since the dawn of the Space Age. In modern times, the increasing spatial, spectral and temporal resolution of Earth-observation satellites has intensified the imbalance between the massive volume of imagery generated onboard and the limited downlink capacity available. In order to overcome this bottleneck, this study analyses and evaluates image compression techniques that enable more efficient use of satellite communication channels. After a throughout review of the main communication and onboard constraints, as well as the most relevant compression paradigms and standards, a set of representative algorithms is experimentally evaluated in two contrasting mission scenarios: a optical CubeSat in low-earth orbit and amid-clas hyperspectral satellite . Results show that no single codec is universally optimal: predictive methods offer low complexity for resource-constrained platforms, while transform-based approaches provide superior rate-distortion performance in high-fidelity missions. The analysis further indicates that relying solely on image quality metrics is inadequate for onboard evaluation. Factors such as compression ratio, execution time, and peak memory usage should be evaluated together, as they collectively influence the practicality of implementation on real hardware. The study highlights the need for mission-driven codec selection, balancing compression efficiency, image quality and onboard resource usage to maximise the scientific return of satellite data. Lightweight machine-learning-based compression as well as ROI-driven or event-based strategies and bandwidth-aware coding or join source-channel schemes are key for future work and development of new solutions for Earth-observation satellites.

© CBLTIC Campus del Baix Llobregat - UPC