Binary steering in discrete tomography reconstruction with sequential and simultaneous iterative algorithms

Research output: Contribution to journalArticlepeer-review

Abstract

The binary steering process is a heuristic designed to intervene between consecutive steps of a nonbinary iterative image reconstruction algorithm in order to gradually steer the iterates towards a binary solution. We present computational results which show that a strongly overrelaxed simultaneous nonbinary iterative algorithm performs in our experiments better than a strongly underrelaxed sequential iterative algorithm. We also notice that faster binary steering gives better binary reconstructed images when the sequential iterative nonbinary algorithm is used.

Original languageEnglish
Pages (from-to)111-124
Number of pages14
JournalLinear Algebra and Its Applications
Volume339
Issue number1-3
DOIs
StatePublished - 15 Dec 2001

Bibliographical note

Funding Information:
We thank Samuel Matej for his collaboration on the first report on binary steering [5], and Gabor Herman and Attila Kuba for many useful discussions. We are indebted to Ronen Greenspan for his devoted work on the implementation and the computational research. We thank Bruno Carvalho and Gabor Herman for having shared with us their test and reconstructed images of [2] from where we took our test image (called Carvalho above) for our experiments. Two anonymous referees helped us greatly in bringing this paper to its final form and we are grateful to them for their comments. This work was supported by research grants 293/97 and 592/00 from the Israel Science Foundation founded by the Israel Academy of Sciences and Humanities, and by grant HL-28438 of the National Institutes of Health (NIH).

Keywords

  • Binary steering
  • Conflict resolution
  • Discrete tomography

ASJC Scopus subject areas

  • Algebra and Number Theory
  • Numerical Analysis
  • Geometry and Topology
  • Discrete Mathematics and Combinatorics

Fingerprint

Dive into the research topics of 'Binary steering in discrete tomography reconstruction with sequential and simultaneous iterative algorithms'. Together they form a unique fingerprint.

Cite this