A multiphase logic framework for multichannel image segmentation

Matthew S. Keegan, Berta Sandberg, Tony F. Chan

Research output: Contribution to journalArticlepeer-review

10 Scopus citations

Abstract

We propose a novel framework for energy-based multiphase segmentation over multiple channels. The framework allows the user to combine the information from each channel as the user sees fit, and thus allows the user to define how the information from each channel should inuence the result. The framework extends the two-phase Logic Framework [J. Vis. Commun. Image R. 16 (2005) 333-358] model. The logic operators of the Logic Framework are used to define objective functions for multiple phases and a condition is defined that prevents conict between energy terms. This condition prevents local minima that may occur using ad hoc methods, such as summing the objective functions of each region.

Original languageEnglish (US)
Pages (from-to)95-110
Number of pages16
JournalInverse Problems and Imaging
Volume6
Issue number1
DOIs
StatePublished - Feb 2012
Externally publishedYes

Keywords

  • Image segmentation

ASJC Scopus subject areas

  • Control and Optimization
  • Analysis
  • Discrete Mathematics and Combinatorics
  • Modeling and Simulation

Fingerprint

Dive into the research topics of 'A multiphase logic framework for multichannel image segmentation'. Together they form a unique fingerprint.

Cite this