Douglas Hope, Hope Scientific Renaissance LLC; Stuart Jefferies, Georgia State University
Keywords: Imaging through Turbulence, Image Restoration
Abstract:
Blind deconvolution (BD) is a commonly used image restoration technique in the ground-based surveillance of space. In addition to providing a blur-free estimate of the object under investigation, BD also provides an estimate of the system point-spread function (PSF). The latter is typically modeled using a Fourier Optics description of the PSF in which the wave front phase is a variable in the restoration. However, there is no guarantee that the estimated wave front is physically realistic which consequently limits the fidelity of the recovered PSF and thus also the recovered object. In this research our hypothesis is that the temporal fluctuations in the Fourier spectra of the recorded image data encode information about the evolution of the wave front. By leveraging these temporal correlations BD should be able to provide realistic estimates of the wave front phase, and a more faithful estimate of the target. It should also increase the robustness of BD of imagery acquired through atmospheric turbulence.
While knowledge of the wave-front phase is most important, many imaging scenarios exist where perturbations in the wave-front amplitude (e.g., near horizon imaging or horizontal imaging) contribute to degradation of the image quality. In these cases, we will demonstrate how modeling the temporal fluctuations in the Fourier spectra of the image data may lead to estimates of the complex wave-front (amplitude and phase), thus further improving the quality of BD restorations.
Date of Conference: September 11-14, 2018
Track: Adaptive Optics & Imaging