Author

Noah Block

Abstract

Sparse-aperture telescopes are a promising technology to increase the resolution of a telescope without the costs of creating very large optics. The individual subapertures in a sparse-aperture system capture a portion of the incoming wavefront to synthesize a monolithic telescope with a larger diameter. This allows higher resolutions to be recorded than any of the individual subapertures are capable of. The resulting image quality of a sparse-aperture system is poorer than obtained from a Cassegrain system that has the same effective diameter. Published research to date on sparse-aperture systems has focused on panchromatic imaging using a “gray-world” model where the input is a monochrome image that is convolved with a specific point spread function. This research uses the spectral sparse-aperture model Robert Introne created for his Ph.D. dissertation to compare the image quality of the polychromatic and gray-world models. Introne’s model creates both the polychromatic and gray-world scenarios. The first experiment performed compares the Golay-6 and triarm configurations for both the polychromatic and gray-world models. The second experiment calculates the threshold when spectral artifacts become apparent and how they evolve in a restored image for the triarm configuration. This is achieved by increasing the amount of introduced phase error in small increments. The behavior of the spectral artifacts can be observed for each scene via this method. The next step in the research is the modeling of a multispectral (MS) sparse-aperture system. This attempts to reduce the effect of spectrally induced artifacts by capturing multiple bands and restoring each one separately, then summing the bands into a panchromatic image to increase the signal-to-noise ratio (SNR). The restorations for the previous experiments all use a scenario where the phase error introduced into the pupil function is perfectly known for the restoration filter. This is not a realistic scenario. An error analysis based on the expected performance of a phase retrieval algorithm is employed to estimate the phase error of the true pupil function. The estimated pupil function is then used to restore the scene degraded by the true pupil function.

Library of Congress Subject Headings

Telescopes; Imaging systems--Image quality; Optics

Publication Date

9-1-2005

Document Type

Thesis

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

Easton, Roger

Advisor/Committee Member

Fienup, James

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: QB88 .B56 2005

Campus

RIT – Main Campus

Share

COinS