Projected barzilai-borwein method with infeasible iterates for nonnegative least-squares image deblurring

Kathleen Fraser, Dirk V. Arnold, Graham Dellaire

Résultat de recherche: Conference contribution

2 Citations (Scopus)

Résumé

We present a non-monotonic gradient descent algorithm with infeasible iterates for the nonnegatively constrained least-squares deblurring of images. The skewness of the intensity values of the deblurred image is used to establish a criterion for when to enforce the nonnegativity constraints. The approach is observed on several test images to either perform comparably to or to outperform a non-monotonic gradient descent approach that does not use infeasible iterates, as well as the gradient projected conjugate gradients algorithm. Our approach is distinguished from the latter by lower memory requirements, making it suitable for use with large, three-dimensional images common in medical imaging.

Langue d'origineEnglish
Titre de la publication principaleProceedings - Conference on Computer and Robot Vision, CRV 2014
Maison d'éditionIEEE Computer Society
Pages189-194
Nombre de pages6
ISBN (imprimé)9781479943388
DOI
Statut de publicationPublished - 2014
Événement11th Conference on Computer and Robot Vision, CRV 2014 - Montreal, QC, Canada
Durée: mai 6 2014mai 9 2014

Séries de publication

PrénomProceedings - Conference on Computer and Robot Vision, CRV 2014

Conference

Conference11th Conference on Computer and Robot Vision, CRV 2014
Pays/TerritoireCanada
VilleMontreal, QC
Période5/6/145/9/14

ASJC Scopus Subject Areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition

Empreinte numérique

Plonger dans les sujets de recherche 'Projected barzilai-borwein method with infeasible iterates for nonnegative least-squares image deblurring'. Ensemble, ils forment une empreinte numérique unique.

Citer