Combining hopfield neural network and contouring methods to enhance super-resolution mapping

Yuan Fong Su, Giles M. Foody, Anuar Mikdad Muad, Ke Sheng Cheng

Research output: Contribution to journalArticle

41 Citations (Scopus)

Abstract

The mixed pixel problem may be reduced through the use of a soft image classification and super-resolution mapping analyses. Here, the positive attributes of two popular super-resolution mapping methods, based on contouring and the Hopfield neural network, are combined. For a binary classification scenario, the method is based on fitting a contour of equal class membership to a pre-final output of a standard Hopfield neural network. Analyses of simulated and real image data sets show that the proposed method is more accurate than the standard contouring and Hopfield neural network based methods, with error typically reduced by a factor of two or more. The sensitivity of the Hopfield neural network based approaches to the setting of a gain function is also explored.

Original languageEnglish
Article number6353566
Pages (from-to)1403-1417
Number of pages15
JournalIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Volume5
Issue number5
DOIs
Publication statusPublished - 2012

Fingerprint

Hopfield neural networks
mapping method
image classification
Image classification
pixel
Pixels
method

Keywords

  • Contour-based approach
  • Hopfield neural network
  • soft classification
  • super-resolution mapping

ASJC Scopus subject areas

  • Computers in Earth Sciences
  • Atmospheric Science

Cite this

Combining hopfield neural network and contouring methods to enhance super-resolution mapping. / Su, Yuan Fong; Foody, Giles M.; Muad, Anuar Mikdad; Cheng, Ke Sheng.

In: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, Vol. 5, No. 5, 6353566, 2012, p. 1403-1417.

Research output: Contribution to journalArticle

@article{bd5d54fdd0f442059fa1e4f3dce64fa8,
title = "Combining hopfield neural network and contouring methods to enhance super-resolution mapping",
abstract = "The mixed pixel problem may be reduced through the use of a soft image classification and super-resolution mapping analyses. Here, the positive attributes of two popular super-resolution mapping methods, based on contouring and the Hopfield neural network, are combined. For a binary classification scenario, the method is based on fitting a contour of equal class membership to a pre-final output of a standard Hopfield neural network. Analyses of simulated and real image data sets show that the proposed method is more accurate than the standard contouring and Hopfield neural network based methods, with error typically reduced by a factor of two or more. The sensitivity of the Hopfield neural network based approaches to the setting of a gain function is also explored.",
keywords = "Contour-based approach, Hopfield neural network, soft classification, super-resolution mapping",
author = "Su, {Yuan Fong} and Foody, {Giles M.} and Muad, {Anuar Mikdad} and Cheng, {Ke Sheng}",
year = "2012",
doi = "10.1109/JSTARS.2012.2191537",
language = "English",
volume = "5",
pages = "1403--1417",
journal = "IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing",
issn = "1939-1404",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "5",

}

TY - JOUR

T1 - Combining hopfield neural network and contouring methods to enhance super-resolution mapping

AU - Su, Yuan Fong

AU - Foody, Giles M.

AU - Muad, Anuar Mikdad

AU - Cheng, Ke Sheng

PY - 2012

Y1 - 2012

N2 - The mixed pixel problem may be reduced through the use of a soft image classification and super-resolution mapping analyses. Here, the positive attributes of two popular super-resolution mapping methods, based on contouring and the Hopfield neural network, are combined. For a binary classification scenario, the method is based on fitting a contour of equal class membership to a pre-final output of a standard Hopfield neural network. Analyses of simulated and real image data sets show that the proposed method is more accurate than the standard contouring and Hopfield neural network based methods, with error typically reduced by a factor of two or more. The sensitivity of the Hopfield neural network based approaches to the setting of a gain function is also explored.

AB - The mixed pixel problem may be reduced through the use of a soft image classification and super-resolution mapping analyses. Here, the positive attributes of two popular super-resolution mapping methods, based on contouring and the Hopfield neural network, are combined. For a binary classification scenario, the method is based on fitting a contour of equal class membership to a pre-final output of a standard Hopfield neural network. Analyses of simulated and real image data sets show that the proposed method is more accurate than the standard contouring and Hopfield neural network based methods, with error typically reduced by a factor of two or more. The sensitivity of the Hopfield neural network based approaches to the setting of a gain function is also explored.

KW - Contour-based approach

KW - Hopfield neural network

KW - soft classification

KW - super-resolution mapping

UR - http://www.scopus.com/inward/record.url?scp=84869494118&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84869494118&partnerID=8YFLogxK

U2 - 10.1109/JSTARS.2012.2191537

DO - 10.1109/JSTARS.2012.2191537

M3 - Article

AN - SCOPUS:84869494118

VL - 5

SP - 1403

EP - 1417

JO - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

JF - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

SN - 1939-1404

IS - 5

M1 - 6353566

ER -