guest ::
| Home > Conference materials > Papers > Hyperspectral image segmentation for estimation of biomass at reclaimed heaps |
Original title:
Hyperspectral image segmentation for estimation of biomass at reclaimed heaps
Authors: Pikl, Miroslav ; Zemek, František
Document type: Papers
Conference/Event: Global Change and Resilience, Brno (CZ), 2013-05-22 / 2013-05-24
Year: 2013
Language: eng
Abstract: This paper presents the preliminary results from a study that aims at estimation of above ground biomass and soil carbon content at reclaimed mining heaps in the Sokolov region. Two image segmentation methods are presented. We applied maximal likelihood (ML) and neural network (NN) classifi ers on airborne hyperspectral data. Th e objective of this part of the study was to prepare a land cover classifi cation of the region. Th e main focus was paid to discrimination of six classes with prevailing forest species cover. Th e classifi cation accuracy of the training sites was 93.75 % for NN and 79.12 % for ML respectively. But ML outperformed NN in overall classifi cation accuracy with 61.54 % compared to 40.9 % of NN. Th e more accurate results of the ML classifi er are probably infl uenced by properties of the training samples. Th e larger size of the training samples derived for ML enabled better representation of class histograms. Th e lower overall NN accuracy could result from high spatial resolution of HS data.
Keywords: classification; hyperspectral; maximal likehood; neural network
Project no.: ED1.1.00/02.0073, OC09001 (CEP), LM2010007 (CEP)
Funding provider: GA MŠk, GA MŠk, GA MŠk
Host item entry: Global Change and Resilience: From Impacts to Responses : Proceedings of the 3rd annual Global Change and Resilience Conference, ISBN 978-80-904351-8-6
Institution: Global Change Research Institute AS ČR (web)
Document availability information: Fulltext is available at the institute of the Academy of Sciences.
Original record: http://hdl.handle.net/11104/0230119
Permalink: http://www.nusl.cz/ntk/nusl-170363
The record appears in these collections:
Research > Institutes ASCR > Global Change Research Institute
Conference materials > Papers
Authors: Pikl, Miroslav ; Zemek, František
Document type: Papers
Conference/Event: Global Change and Resilience, Brno (CZ), 2013-05-22 / 2013-05-24
Year: 2013
Language: eng
Abstract: This paper presents the preliminary results from a study that aims at estimation of above ground biomass and soil carbon content at reclaimed mining heaps in the Sokolov region. Two image segmentation methods are presented. We applied maximal likelihood (ML) and neural network (NN) classifi ers on airborne hyperspectral data. Th e objective of this part of the study was to prepare a land cover classifi cation of the region. Th e main focus was paid to discrimination of six classes with prevailing forest species cover. Th e classifi cation accuracy of the training sites was 93.75 % for NN and 79.12 % for ML respectively. But ML outperformed NN in overall classifi cation accuracy with 61.54 % compared to 40.9 % of NN. Th e more accurate results of the ML classifi er are probably infl uenced by properties of the training samples. Th e larger size of the training samples derived for ML enabled better representation of class histograms. Th e lower overall NN accuracy could result from high spatial resolution of HS data.
Keywords: classification; hyperspectral; maximal likehood; neural network
Project no.: ED1.1.00/02.0073, OC09001 (CEP), LM2010007 (CEP)
Funding provider: GA MŠk, GA MŠk, GA MŠk
Host item entry: Global Change and Resilience: From Impacts to Responses : Proceedings of the 3rd annual Global Change and Resilience Conference, ISBN 978-80-904351-8-6
Institution: Global Change Research Institute AS ČR (web)
Document availability information: Fulltext is available at the institute of the Academy of Sciences.
Original record: http://hdl.handle.net/11104/0230119
Permalink: http://www.nusl.cz/ntk/nusl-170363
The record appears in these collections:
Research > Institutes ASCR > Global Change Research Institute
Conference materials > Papers
Record created 2014-01-30, last modified 2021-11-24