Reservoir Computing for Prediction of the Spatially-Variant Point Spread Function
Type of content
Journal Article
UC permalink
Publisher's DOI/URI
Thesis discipline
Degree name
Publisher
University of Canterbury. Electrical and Computer Engineering
Journal Title
Journal ISSN
Volume Title
Language
Date
2008
Authors
Weddell, S.J.
Webb, R.Y.
Abstract
A new method is presented which provides prediction of the spatially variant point spread function for the restoration of astronomical images, distorted by atmospheric turbulence when viewed using ground-based telescopes. Our approach uses reservoir computing to firstly learn the spatio-temporal evolution of aberrations caused by turbulence, and secondly, predicts the space-varying point spread function (PSF) for application of widely-used deconvolution algorithms, resulting in the restoration of astronomical images. In this article, a reservoir-based, recurrent neural network is used to predict modal aberrations that comprise the spatially variant PSF over a wide field-of-view using a time-series ensemble from multiple reference beacons.
Description
Citation
Weddell, S.J., Webb, R.Y. (2008) Reservoir Computing for Prediction of the Spatially-Variant Point Spread Function. IEEE Journal of Selected Topics in Signal Processing, 2(5), pp. 624-634.
Keywords
Signal Processing, Adaptive Optics, Neural Networks, Fields of Research::240000 Physical Sciences::240100 Astronomical Sciences::240199 Astronomical sciences not elsewhere classified