A Group-Lasso Active Set Strategy for Multiclass Hyperspectral Image Classification
Hyperspectral images have a strong potential for landcover/landuse classification, since the spectra of the pixels can highlight subtle differences between materials and provide information beyond the visible spectrum. Yet, a limitation of most current approaches is the hypothesis of spatial indepen...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Copernicus Publications
2014-08-01
|
Series: | ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
Online Access: | http://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/II-3/157/2014/isprsannals-II-3-157-2014.pdf |
_version_ | 1811295293930471424 |
---|---|
author | D. Tuia N. Courty R. Flamary |
author_facet | D. Tuia N. Courty R. Flamary |
author_sort | D. Tuia |
collection | DOAJ |
description | Hyperspectral images have a strong potential for landcover/landuse classification, since the spectra of the pixels can highlight subtle
differences between materials and provide information beyond the visible spectrum. Yet, a limitation of most current approaches is
the hypothesis of spatial independence between samples: images are spatially correlated and the classification map should exhibit
spatial regularity. One way of integrating spatial smoothness is to augment the input spectral space with filtered versions of the bands.
However, open questions remain, such as the selection of the bands to be filtered, or the filterbank to be used. In this paper, we consider
the entirety of the possible spatial filters by using an incremental feature learning strategy that assesses whether a candidate feature
would improve the model if added to the current input space. Our approach is based on a multiclass logistic classifier with group-lasso
regularization. The optimization of this classifier yields an optimality condition, that can easily be used to assess the interest of a
candidate feature without retraining the model, thus allowing drastic savings in computational time. We apply the proposed method
to three challenging hyperspectral classification scenarios, including agricultural and urban data, and study both the ability of the
incremental setting to learn features that always improve the model and the nature of the features selected. |
first_indexed | 2024-04-13T05:31:08Z |
format | Article |
id | doaj.art-5a1f24c3ab6e4706ac7f5ad12847d0fa |
institution | Directory Open Access Journal |
issn | 2194-9042 2194-9050 |
language | English |
last_indexed | 2024-04-13T05:31:08Z |
publishDate | 2014-08-01 |
publisher | Copernicus Publications |
record_format | Article |
series | ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
spelling | doaj.art-5a1f24c3ab6e4706ac7f5ad12847d0fa2022-12-22T03:00:26ZengCopernicus PublicationsISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences2194-90422194-90502014-08-01II-315716410.5194/isprsannals-II-3-157-2014A Group-Lasso Active Set Strategy for Multiclass Hyperspectral Image ClassificationD. Tuia0N. Courty1R. Flamary2EPFL, Laboratory of Geographic Information Systems, Lausanne, SwitzerlandUniversité de Bretagne du Sud, IRISA, Vannes, FranceUniversité de Nice Sophia-Antipolis, Lab. Lagrange, UMR CNRS 7293, Nice, FranceHyperspectral images have a strong potential for landcover/landuse classification, since the spectra of the pixels can highlight subtle differences between materials and provide information beyond the visible spectrum. Yet, a limitation of most current approaches is the hypothesis of spatial independence between samples: images are spatially correlated and the classification map should exhibit spatial regularity. One way of integrating spatial smoothness is to augment the input spectral space with filtered versions of the bands. However, open questions remain, such as the selection of the bands to be filtered, or the filterbank to be used. In this paper, we consider the entirety of the possible spatial filters by using an incremental feature learning strategy that assesses whether a candidate feature would improve the model if added to the current input space. Our approach is based on a multiclass logistic classifier with group-lasso regularization. The optimization of this classifier yields an optimality condition, that can easily be used to assess the interest of a candidate feature without retraining the model, thus allowing drastic savings in computational time. We apply the proposed method to three challenging hyperspectral classification scenarios, including agricultural and urban data, and study both the ability of the incremental setting to learn features that always improve the model and the nature of the features selected.http://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/II-3/157/2014/isprsannals-II-3-157-2014.pdf |
spellingShingle | D. Tuia N. Courty R. Flamary A Group-Lasso Active Set Strategy for Multiclass Hyperspectral Image Classification ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
title | A Group-Lasso Active Set Strategy for Multiclass Hyperspectral Image Classification |
title_full | A Group-Lasso Active Set Strategy for Multiclass Hyperspectral Image Classification |
title_fullStr | A Group-Lasso Active Set Strategy for Multiclass Hyperspectral Image Classification |
title_full_unstemmed | A Group-Lasso Active Set Strategy for Multiclass Hyperspectral Image Classification |
title_short | A Group-Lasso Active Set Strategy for Multiclass Hyperspectral Image Classification |
title_sort | group lasso active set strategy for multiclass hyperspectral image classification |
url | http://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/II-3/157/2014/isprsannals-II-3-157-2014.pdf |
work_keys_str_mv | AT dtuia agrouplassoactivesetstrategyformulticlasshyperspectralimageclassification AT ncourty agrouplassoactivesetstrategyformulticlasshyperspectralimageclassification AT rflamary agrouplassoactivesetstrategyformulticlasshyperspectralimageclassification AT dtuia grouplassoactivesetstrategyformulticlasshyperspectralimageclassification AT ncourty grouplassoactivesetstrategyformulticlasshyperspectralimageclassification AT rflamary grouplassoactivesetstrategyformulticlasshyperspectralimageclassification |