A sure-fired way to choose smoothing parameters in ill-conditioned inverse problems
Author:
- 23 December 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 3, 89-92
Abstract
Regularisation methods for the solution of inverse problems are well known although the theoretical study of their performance especially in image processing contexts is not well advanced. What is also much less resolved is smoothing or penalty parameter estimation. We describe a general procedure for estimation of auxiliary finite dimensional parameters in ill-conditioned inverse problems. The method is applicable to nonlinear problems, involves no approximations but offers computational advantages over cross validation and maximum likelihood.Keywords
This publication has 11 references indexed in Scilit:
- Approximate maximum likelihood hyperparameter estimation for Gibbs priors2002
- On some Bayesian choices of regularization parameter in image restorationInverse Problems, 1993
- A generalized Gaussian image model for edge-preserving MAP estimationIEEE Transactions on Image Processing, 1993
- Data-dependent bandwidth selection for emission computed tomography reconstructionIEEE Transactions on Medical Imaging, 1993
- Methods for choosing the regularization parameter and estimating the noise variance in image restoration and their relationIEEE Transactions on Image Processing, 1992
- Spline Models for Observational Data1990
- Density Estimation for Statistics and Data Analysis1986
- From Stein's Unbiased Risk Estimates to the Method of Generalized Cross ValidationThe Annals of Statistics, 1985
- Estimation of the Mean of a Multivariate Normal DistributionThe Annals of Statistics, 1981
- Some Comments on C PTechnometrics, 1973