New termination rule for linear iterative image restoration algorithms

The paper addresses primarily the monitoring of iterative methods for image restoration by introducing a new technique for measuring algorithm performance. A filtering analysis relating the solution computed by a particular method to the solution space as described by the singular value decompositio...

Full description

Saved in:
Bibliographic Details
Published in:Optical Engineering 1990-05, Vol.29 (5), p.471-477
Main Authors: Katsaggelos, Aggelos K, Sullivan, Barry J
Format: Article
Language:English
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The paper addresses primarily the monitoring of iterative methods for image restoration by introducing a new technique for measuring algorithm performance. A filtering analysis relating the solution computed by a particular method to the solution space as described by the singular value decomposition of the system matrix is considered. Generation of filter functions by two iterative algorithms - steepest descent and conjugate gradients - is demonstrated; these functions are used to compute the distance between the desired image and the restored image at each iteration. An iterative framework for the image restoration problem is presented, along with a demonstration by example of the use of one of the distance measures presented as a termination rule. (V.T.)
ISSN:0091-3286
1560-2303
DOI:10.1117/12.55615