Land cover mapping at the sub-pixel scale using a Hopfield neural network
Land cover mapping at the sub-pixel scale using a Hopfield neural network
Fuzzy classification techniques have been developed to estimate the class composition of image pixels, but their output provides no indication of how these classes are distributed spatially within the pixel. As such, it remains for robust techniques that provide improved spatial representation of land cover to be developed. The use of a Hopfield neural network to map the spatial distribution of classes more reliably was investigated. An approach was adopted which used the output from a fuzzy classification to constrain a Hopfield neural network formulated as an energy minimisation tool. The energy minimum represents a ‘bestguess’ map of the spatial distribution of class components in each pixel. The technique was applied to both synthetic and Landsat TM imagery. The resultant maps provided an accurate and improved representation of the land covers studied. The Hopfield neural network used in this way represents a simple, robust and efficient tool for mapping land cover from remotely sensed imagery at the sub-pixel scale.
47-51
International Symposium on Remote Sensing of the Environment
Tatem, Andrew J.
6c6de104-a5f9-46e0-bb93-a1a7c980513e
Lewis, Hugh G.
e9048cd8-c188-49cb-8e2a-45f6b316336a
Atkinson, Peter
96e96579-56fe-424d-a21c-17b6eed13b0b
Nixon, Mark S.
2b5b9804-5a81-462a-82e6-92ee5fa74e12
March 2000
Tatem, Andrew J.
6c6de104-a5f9-46e0-bb93-a1a7c980513e
Lewis, Hugh G.
e9048cd8-c188-49cb-8e2a-45f6b316336a
Atkinson, Peter
96e96579-56fe-424d-a21c-17b6eed13b0b
Nixon, Mark S.
2b5b9804-5a81-462a-82e6-92ee5fa74e12
Tatem, Andrew J., Lewis, Hugh G., Atkinson, Peter and Nixon, Mark S.
(2000)
Land cover mapping at the sub-pixel scale using a Hopfield neural network.
In Proceedings of the 28th International Symposium on Remote Sensing of the Environment.
International Symposium on Remote Sensing of the Environment.
.
Record type:
Conference or Workshop Item
(Paper)
Abstract
Fuzzy classification techniques have been developed to estimate the class composition of image pixels, but their output provides no indication of how these classes are distributed spatially within the pixel. As such, it remains for robust techniques that provide improved spatial representation of land cover to be developed. The use of a Hopfield neural network to map the spatial distribution of classes more reliably was investigated. An approach was adopted which used the output from a fuzzy classification to constrain a Hopfield neural network formulated as an energy minimisation tool. The energy minimum represents a ‘bestguess’ map of the spatial distribution of class components in each pixel. The technique was applied to both synthetic and Landsat TM imagery. The resultant maps provided an accurate and improved representation of the land covers studied. The Hopfield neural network used in this way represents a simple, robust and efficient tool for mapping land cover from remotely sensed imagery at the sub-pixel scale.
This record has no associated files available for download.
More information
Published date: March 2000
Additional Information:
Organisation: CSIR, NASA, SAI, ESA...
Venue - Dates:
28th International Symposium on Remote Sensing of Environment, Cape Town, South Africa, Cape Town, South Africa, 2000-03-27 - 2000-03-31
Organisations:
Southampton Wireless Group
Identifiers
Local EPrints ID: 252956
URI: http://eprints.soton.ac.uk/id/eprint/252956
PURE UUID: c0879901-c7d8-425e-bc40-a97960764423
Catalogue record
Date deposited: 11 Apr 2000
Last modified: 23 Feb 2023 02:59
Export record
Contributors
Author:
Peter Atkinson
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics