Subpixel multiframe registration for formation flying spacecraft

This thesis develops a multiframe image registration algorithm to accurately estimate the motion parameters of a sequence of images with constant linear translation between frames. The algorithm is developed with formation flying applications in mind, where such motion is common. The algorithm is non-iterative (obtaining a registration estimate in a fixed amount of time), efficient (requiring only FFT and spatial scaling operations), and parameterless (requiring no tuning for image classes). Additionally, the algorithm performs well under extreme levels of noise, obtaining an accurate motion estimate even when individual frames are so severely degraded that high-frequency structures are no longer visible in the individual frames. The algorithm is tested against synthetic noisy frames which have been generated by a computational pipeline designed to simulate observations that will be made by the VIrtual Super-resolution Optics with Reconfigurable Swarms (VISORS) Cubesat mission, set to be launched in 2023.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Electrical & Computer Engr

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2021

Author dc:creator

• Widloski, Evan

Contributors dc:contributor

• Kamalabadi, Farzad

• image registration
• signal processing
• remote sensing
• formation flying

Statement dc:rights

• Copyright 2020 Evan Widloski

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/109458

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/109458