Evaluation of classification methods with polarimetric ALOS/PALSAR data

Anne Lönnqvist; Yrjö Rauste; Heikki Ahola; Matthieu Molinier; Tuomas Häme

Evaluation of classification methods with polarimetric ALOS/PALSAR data

Anne Lönnqvist, Yrjö Rauste, Heikki Ahola, Matthieu Molinier, Tuomas Häme

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific

Abstract

This article compares four land cover classification methods using ALOS Palsar fully polarimetric data. Two methods are based on supervised classification and two on unsupervised classification. Both full polarimetric and intensity data were used. Fully polarimetric data gave better results (85.5%–80.8%) than intensity data only (84%–76.1%) but the differences in the overall accuracies between the methods were not more than 6%. When using intensity data only, classification results with HH and HV features were as good as using in additionally also VV, HH+VV and HH-VV bands.

Original language	English
Title of host publication	Forests and Remote Sensing
Subtitle of host publication	Methods and Operational Tools (Forestsat 2007)
Number of pages	5
Publication status	Published - 2007
MoE publication type	B3 Non-refereed article in conference proceedings
Event	Forests and Remote Sensing: Methods and Operational Tools, Forestsat 2007 - Montpellier, France Duration: 5 Nov 2007 → 7 Nov 2007

Conference

Conference	Forests and Remote Sensing: Methods and Operational Tools, Forestsat 2007
Abbreviated title	Forestsat
Country/Territory	France
City	Montpellier
Period	5/11/07 → 7/11/07

Keywords

Polarimetric synthetic aperture radar
land cover

Access to Document

https://publications.vtt.fi/julkaisut/muut/2007/Forestsat_2007.pdf

Cite this

@inproceedings{5c7ffa5e380c4c1984be00654dfadffb,

title = "Evaluation of classification methods with polarimetric ALOS/PALSAR data",

abstract = "This article compares four land cover classification methods using ALOS Palsar fully polarimetric data. Two methods are based on supervised classification and two on unsupervised classification. Both full polarimetric and intensity data were used. Fully polarimetric data gave better results (85.5%–80.8%) than intensity data only (84%–76.1%) but the differences in the overall accuracies between the methods were not more than 6%. When using intensity data only, classification results with HH and HV features were as good as using in additionally also VV, HH+VV and HH-VV bands.",

keywords = "Polarimetric synthetic aperture radar, land cover",

author = "Anne L{\"o}nnqvist and Yrj{\"o} Rauste and Heikki Ahola and Matthieu Molinier and Tuomas H{\"a}me",

note = "Poster presentation, CD proceedings for participants only Projektin nimi NewSAR Project code: 7859 ; Forests and Remote Sensing: Methods and Operational Tools, Forestsat 2007, Forestsat ; Conference date: 05-11-2007 Through 07-11-2007",

year = "2007",

language = "English",

booktitle = "Forests and Remote Sensing",

}

Lönnqvist, A, Rauste, Y, Ahola, H, Molinier, M & Häme, T 2007, Evaluation of classification methods with polarimetric ALOS/PALSAR data. in Forests and Remote Sensing: Methods and Operational Tools (Forestsat 2007). Forests and Remote Sensing: Methods and Operational Tools, Forestsat 2007, Montpellier, France, 5/11/07. <https://publications.vtt.fi/julkaisut/muut/2007/Forestsat_2007.pdf>

TY - GEN

T1 - Evaluation of classification methods with polarimetric ALOS/PALSAR data

AU - Lönnqvist, Anne

AU - Rauste, Yrjö

AU - Ahola, Heikki

AU - Molinier, Matthieu

AU - Häme, Tuomas

N1 - Poster presentation, CD proceedings for participants only Projektin nimi NewSAR Project code: 7859

PY - 2007

Y1 - 2007

N2 - This article compares four land cover classification methods using ALOS Palsar fully polarimetric data. Two methods are based on supervised classification and two on unsupervised classification. Both full polarimetric and intensity data were used. Fully polarimetric data gave better results (85.5%–80.8%) than intensity data only (84%–76.1%) but the differences in the overall accuracies between the methods were not more than 6%. When using intensity data only, classification results with HH and HV features were as good as using in additionally also VV, HH+VV and HH-VV bands.

AB - This article compares four land cover classification methods using ALOS Palsar fully polarimetric data. Two methods are based on supervised classification and two on unsupervised classification. Both full polarimetric and intensity data were used. Fully polarimetric data gave better results (85.5%–80.8%) than intensity data only (84%–76.1%) but the differences in the overall accuracies between the methods were not more than 6%. When using intensity data only, classification results with HH and HV features were as good as using in additionally also VV, HH+VV and HH-VV bands.

KW - Polarimetric synthetic aperture radar

KW - land cover

M3 - Conference article in proceedings

BT - Forests and Remote Sensing

T2 - Forests and Remote Sensing: Methods and Operational Tools, Forestsat 2007

Y2 - 5 November 2007 through 7 November 2007

ER -

Evaluation of classification methods with polarimetric ALOS/PALSAR data

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this