IIT Home Page CNR Home Page

Stopping rules for iterative methods in nonnegatively constrained deconvolution

We consider the two-dimensional discrete nonnegatively constrained  deconvolution problem, whose goal is
to reconstruct an object ${\bf x}^*$ from its image ${\bf b}$ obtained through
an optical system and affected by  noise. When the large size of the problem
prevents regularization by filtering, iterative methods enjoying
semiconvergence property, coupled with suitable strategies for enforcing nonnegativity, are suggested. For these methods an accurate detection of the stopping index is essential. In this paper we analyze various
stopping rules and test their effect on three different iterative regularizing methods, by a large experimentation.


2011

Autori esterni: Grazia Lotti (Dip. di Matematica, University of Parma), Ornella Menchi ( Dip. di Informatica, University of Pisa), Francesco Romani ( Dip. di Informatica, University of Pisa)
Autori IIT:

Tipo: TR Rapporti tecnici
Area di disciplina: Mathematics
IIT TR-15-2011

Attività: Metodi numerici per problemi di grandi dimensioni