Greedy Deconvolution of Point-like Objects - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2009

Greedy Deconvolution of Point-like Objects

Abstract

The orthogonal matching pursuit (OMP) is an algorithm to solve sparse approximation problems. In [1] a sufficient condition for exact recovery is derived, in [2] the authors transfer it to noisy signals. We will use OMP for reconstruction of an inverse problem, namely the deconvolution problem. In sparse approximation problems one often has to deal with the problem of redundancy of a dictionary, i.e. the atoms are not linearly independent. However, one expects them to be approximatively orthogonal and this is quantified by incoherence. This idea cannot be transfered to ill-posed inverse problems since here the atoms are typically far from orthogonal: The illposedness of the (typically compact) operator causes that the correlation of two distinct atoms probably gets huge, i.e. that two atoms can look much alike. Therefore in [3], [4] the authors derive a recovery condition which uses the kind of structure one assumes on the signal and works without the concept of coherence. In this paper we will transfer these results to noisy signals. For our source we assume that it consists of a superposition of point-like objects with an a-priori known distance. We will apply it exemplarily to Dirac peaks convolved with Gaussian kernel as used in mass spectrometry.
Fichier principal
Vignette du fichier
26.pdf (172.32 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

inria-00369571 , version 1 (20-03-2009)

Identifiers

  • HAL Id : inria-00369571 , version 1

Cite

Dirk A. Lorenz, Dennis Trede. Greedy Deconvolution of Point-like Objects. SPARS'09 - Signal Processing with Adaptive Sparse Structured Representations, Inria Rennes - Bretagne Atlantique, Apr 2009, Saint Malo, France. ⟨inria-00369571⟩

Collections

SPARS09
87 View
68 Download

Share

Gmail Facebook X LinkedIn More