Design and analysis of an information theoretic algorithm for vector field segmentation

Ian B. Kerfoot, Yoram Bresler

Research output: Contribution to journalConference article

Abstract

We present several objective functions for vector field segmentation. Leclerc's MRF model is extended by the addition of information-theoretic penalties for regions and distinct means. Massively parallel optimization algorithms are presented. Standard methods of signal detection and estimation are used to develop a theoretical performance analysis, which quantitatively predicts the performance at realistic noise levels. Theoretical and experimental results agree fairly well.

Original languageEnglish (US)
Pages (from-to)92-103
Number of pages12
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume1904
DOIs
StatePublished - Jun 10 1993
EventImage Modeling 1993 - San Jose, United States
Duration: Jan 31 1993Feb 5 1993

Fingerprint

Parallel Optimization
Signal Detection
Signal detection
Parallel Algorithms
Performance Analysis
Penalty
Vector Field
Theoretical Analysis
Optimization Algorithm
Segmentation
Objective function
Distinct
Predict
signal detection
Experimental Results
penalties
optimization
Model
Design
Standards

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

Design and analysis of an information theoretic algorithm for vector field segmentation. / Kerfoot, Ian B.; Bresler, Yoram.

In: Proceedings of SPIE - The International Society for Optical Engineering, Vol. 1904, 10.06.1993, p. 92-103.

Research output: Contribution to journalConference article

@article{2fe8fe5655534470a9bf087102bd944f,
title = "Design and analysis of an information theoretic algorithm for vector field segmentation",
abstract = "We present several objective functions for vector field segmentation. Leclerc's MRF model is extended by the addition of information-theoretic penalties for regions and distinct means. Massively parallel optimization algorithms are presented. Standard methods of signal detection and estimation are used to develop a theoretical performance analysis, which quantitatively predicts the performance at realistic noise levels. Theoretical and experimental results agree fairly well.",
author = "Kerfoot, {Ian B.} and Yoram Bresler",
year = "1993",
month = "6",
day = "10",
doi = "10.1117/12.146682",
language = "English (US)",
volume = "1904",
pages = "92--103",
journal = "Proceedings of SPIE - The International Society for Optical Engineering",
issn = "0277-786X",
publisher = "SPIE",

}

TY - JOUR

T1 - Design and analysis of an information theoretic algorithm for vector field segmentation

AU - Kerfoot, Ian B.

AU - Bresler, Yoram

PY - 1993/6/10

Y1 - 1993/6/10

N2 - We present several objective functions for vector field segmentation. Leclerc's MRF model is extended by the addition of information-theoretic penalties for regions and distinct means. Massively parallel optimization algorithms are presented. Standard methods of signal detection and estimation are used to develop a theoretical performance analysis, which quantitatively predicts the performance at realistic noise levels. Theoretical and experimental results agree fairly well.

AB - We present several objective functions for vector field segmentation. Leclerc's MRF model is extended by the addition of information-theoretic penalties for regions and distinct means. Massively parallel optimization algorithms are presented. Standard methods of signal detection and estimation are used to develop a theoretical performance analysis, which quantitatively predicts the performance at realistic noise levels. Theoretical and experimental results agree fairly well.

UR - http://www.scopus.com/inward/record.url?scp=33749780797&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33749780797&partnerID=8YFLogxK

U2 - 10.1117/12.146682

DO - 10.1117/12.146682

M3 - Conference article

AN - SCOPUS:33749780797

VL - 1904

SP - 92

EP - 103

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

SN - 0277-786X

ER -