Learning optimal spatially-dependent regularization parameters in Total Variation image restoration I
- Authors
- Cao, Van Chung
- Format
- Article
- Status
- publishedVersion
- Description
We consider a bilevel optimization approach in function space for the choice ofspatially dependent regularization parameters in TV image restoration models. First- andsecond-order optimality conditions for the bilevel problem are studied, when the spatially-dependent parameter belongs to the Sobolev space H1(?). A combined Schwarz domaindecomposition-semismooth Newton method is proposed for the solution of the full op-timality system and local superlinear convergence of the semismooth Newton method isanalyzed. Exhaustive numerical computations are ?nally carried out to show the suitabilityof the approach.
Escuela Polit?cnica Nacional
https://www.researchgate.net/publication/301926366_Learning_optimal_spatially-dependent_regularization_parameters_in_total_variation_image_restoration
- Publication Year
- 2016
- Language
- eng
- Topic
- LEARNING OPTIMAL
SPATIALLY-DEPENDENT
IMAGE RESTORATION
- Repository
- Repositorio SENESCYT
- Rights
- openAccess
- License
- openAccess