Deconvolution from wave front sensing using the frozen flow hypothesis

Abstract

Deconvolution from wave front sensing (DWFS) is an image-reconstruction technique for compensating the image degradation due to atmospheric turbulence. DWFS requires the simultaneous recording of high cadence short-exposure images and wave-front sensor (WFS) data. A deconvolution algorithm is then used to estimate both the target object and the wave front phases from the images, subject to constraints imposed by the WFS data and a model of the optical system. Here we show that by capturing the inherent temporal correlations present in the consecutive wave fronts, using the frozen flow hypothesis (FFH) during the modeling, high-quality object estimates may be recovered in much worse conditions than when the correlations are ignored.