Towards real-time management of satellite microvibrations for on-board hyperspectral image quality enhancement

Abstract

The nature of the earth observation satellite systems attitude and orbit control systems, makes them inherently susceptible to microvibrations. The largest sources of microvibrations in the current satellite platforms market are: reaction wheels, mechanical movements inside the thermal control subsystem, the changing thermal environment, and the vibration of the satellite structure due to its flexibility. Microvibrations can be categorized according to its frequency into two classes: low-frequency and high-frequency. The first group affects the positioning accuracy of the image capture system, while the second decreases the image spatial resolution. This work address the second group. The spatial resolution of satellite on-board imaging systems has continuously increased in the last few years. This increase has positioned microvibrations as an important factor in the payload performance budget, many times now becoming a driving limiting factor in the on-space spatial resolution of the earth-observation payloads. This effect is generally complicated to test and calibrate on-ground, due to the difficulties of simulating the real environment at integrated satellite level, and presents more variance on satellite platforms on which the quality control processes are less stringent (like NewSpace).

This work proposes a lightweight multi-image super-resolution algorithm that can help coping with microvibration effects on hyperspectral payloads, resulting in an enhanced spatial and spectral image quality for the same on-board sensor and optics.