Citation

Abstract

The Spectral Airglow Structure Imager (SASI) is a mission on CubeSat NB's VIOLET. Its purpose is to image the 630 nm redline atomic oxygen airglow layer in the ionosphere which is present between the altitudes of 200-300 km. The image acquisition conditions of VIOLET make it susceptible to signicant degradation due to rotational and translational blurring. An image recovery pipeline by ring extraction and linearization using Richardson-Lucy and Wiener methods for egomotion blurring is presented to mitigate its effects. Boundary conditions for pre-processing to handle corner degradation due to the nature of rotational blurring are discussed. Interpolation by means of nearest neighbor, splicing and specic pixel deconvolutions are also discussed to increase image quality after deconvolution. Results show effective deconvolution of both rotationally blurred and egomotion blurred redline ionosphere images, resulting in signicant improvements on Structural Similarity indices (SSIM) across the expected range of blurring for the SASI subsystem.

Acknowledgements of Funding

• the Canadian Space Agency (CSA),
• the New Brunswick Innovation Foundation (NBIF), and
• the University of New Brunswick (UNB).

Thesis