Title: "A Deconvolution Problem in Astronomy"
Speaker: Adriana Gonzalez
Location: "Shannon" Seminar Room (a105) Place du Levant 3, Maxwell Building, 1st floor
Date / Time (duration): Thursday 27/3/2013, 14h00 (~ 45')
Abstract: Optical sensors distort the observation of an object by the imperfection of their elements and by the natural physics of the observation. The acquired image is frequently corrupted by the noise coming from the sensor itself and by the Point Spread Function (PSF) of the instrument, which filters the hypothetically pure image. In most situations, we do not have any knowledge on this PSF, hence we face a blind deconvolution problem when estimating the pure image.
This presentation is dedicated to a specific astronomical application, where some prior information on the image to reconstruct is available. Therefore, the initial blind deconvolu- tion problem can be divided into two different deconvolution sub-problems: (i) estimate the instrument PSF using the a priori information on the image and (ii) recover the pure image using the estimated PSF in (i). Assuming both the PSF and the image have a sparse wavelet representation, the two reconstruction sub-problems can be optimized by promoting a small L1 norm under a data fidelity constraint with the available observations. Moreover, in both sub-problem, the image to reconstruct represents an observation, which is always positive. Therefore, a positivity constraint is added to improve the stability.
Last updated September 25, 2013, at 09:05 AM