Institute of Information Theory and Automation

You are here

Publication details

Fine Structure Recognition in Multichannel Observations

Conference Paper (international conference)

Šimberová Stanislava, Haindl Michal, Šroubek Filip


serial: International Conference on Digital Image Computing Techniques and Applications (DICTA) 2012, p. 1-7

action: International Conference on Digital Image Computing Techniques and Applications (DICTA) 2012, (Fremantle, AU, 03.12.2012-05.12.2012)

project(s): 409/2011, CESNET, GAP103/11/1552, GA ČR, GA102/08/1593, GA ČR, GA102/08/0593, GA ČR

keywords: image restoration, image recognition

preview: Download

abstract (eng):

Two restoration methods applied to the multitemporal solar images are presented. Our main goal is to model and remove degradation in a subimage, where a specific event is investigated. Using information of the input (blurred) channels within a short observed sequence a new undegraded image is reconstructed. Degradation is assumed to follow a linear degradation model with an unknown possibly non-homogeneous point spread function (PSF) and additive noise. The first method ({/bf VAM}) is based on multichannel blind deconvolution (MBD) using a variational approach to blur estimation, while the second one ({/bf SAM}) supposes solution of the multidimensional causal regressive model representing the degraded image (channel). Experimental image data are from the ground based observation (white light) and satellite SOHO mission - EIT (EUV). Contributions of both suggested methods and their generalization are discussed.

RIV: BD

2012-12-21 16:10