Nonsmooth Convex Optimization for Structured Illumination Microscopy Image Reconstruction
Résumé
Our contribution is twofold: first, we investigate the properties of a linear reconstruction method in SIM, discussing its properties from an estimation point of view and its relationship with variational approaches. Second, capitalizing on this formalism , we adopt a variational approach taking into account the type of noise, typically Poissonian, associated to convex nonsmooth regularizing penalties, like the total variation, which have shown their efficiency to solve many imaging inverse problems. We detail the implementation of the proximal algorithm allowing to solve the problems exactly.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...