Fusion of support vector machines for classification of multisensor data

Björn Waske; Jón Atli Benediktsson

doi:10.1109/TGRS.2007.898446

Fusion of support vector machines for classification of multisensor data

Björn Waske
, Jón Atli Benediktsson

University of Iceland

University of Iceland

Research output: Contribution to journal › Article › peer-review

Abstract

The classification of multisensor data sets, consisting of multitemporal synthetic aperture radar data and optical imagery, is addressed. The concept is based on the decision fusion of different outputs. Each data source is treated separately and classified by a support vector machine (SVM). Instead of fusing the final classification outputs (i.e., land cover classes), the original outputs of each SVM discriminant function are used in the subsequent fusion process. This fusion is performed by another SVM, which is trained on the a priori outputs. In addition, two voting schemes are applied to create the final classification results. The results are compared with well-known parametric and nonparametric classifier methods, i.e., decision trees, the maximumlikelihood classifier, and classifier ensembles. The proposed SVM-based fusion approach outperforms all other approaches and significantly improves the results of a single SVM, which is trained on the whole multisensor data set.

Original language	English
Pages (from-to)	3858-3866
Number of pages	9
Journal	IEEE Transactions on Geoscience and Remote Sensing
Volume	45
Issue number	12
DOIs	https://doi.org/10.1109/TGRS.2007.898446
Publication status	Published - Dec 2007

Bibliographical note

Funding Information: B. Waske visited the Department of Electrical and Computer Engineering, University of Iceland, funded by the ENVILAND research project. He would like to thank G. Menz, M. Braun, and V. Heinzel from the University of Bonn. The software for the SVM classification was developed by A. Janz and S. Schiefer, Humboldt-Universität, Berlin, Germany. The authors would like to thank the European Space Agency for providing Envisat ASAR and ERS-2 data through a CAT 1 proposal (C1.3115). The SPOT image is provided through the European OASIS program (OASIS 58 - CE 6324). The data were acquired within the ENVILAND research project (FKZ 50EE0404), funded by the German Aerospace Center (DLR) and the Federal Ministry of Economics and Technology (BMWi). The authors would also like to thank the comments of the anonymous reviewers who helped us significantly improve this paper. Funding Information: Manuscript received December 10, 2006; revised March 19, 2007. This work was supported in part by the German Aerospace Center (DLR) and Federal Ministry of Economics and Technology (BMWi) under the project Enviland (FKZ 50EE0404), as well as an ESA Cat-1 proposals (C1P 3115) and European OASIS program (OASIS 58—CE 6324) and in part by the German Research Foundation (DFG) under the Research Training Group 722 (Information Techniques for Precision Crop Protection) at the University of Bonn.

Other keywords

Data fusion
Multisensor imagery
Multispectral data
Support vector machines (SVM)
Synthetic aperture radar (SAR) data

Access to Document

10.1109/TGRS.2007.898446

Cite this

@article{5aa8456eb7d5454c94584d0657f20b78,

title = "Fusion of support vector machines for classification of multisensor data",

abstract = "The classification of multisensor data sets, consisting of multitemporal synthetic aperture radar data and optical imagery, is addressed. The concept is based on the decision fusion of different outputs. Each data source is treated separately and classified by a support vector machine (SVM). Instead of fusing the final classification outputs (i.e., land cover classes), the original outputs of each SVM discriminant function are used in the subsequent fusion process. This fusion is performed by another SVM, which is trained on the a priori outputs. In addition, two voting schemes are applied to create the final classification results. The results are compared with well-known parametric and nonparametric classifier methods, i.e., decision trees, the maximumlikelihood classifier, and classifier ensembles. The proposed SVM-based fusion approach outperforms all other approaches and significantly improves the results of a single SVM, which is trained on the whole multisensor data set.",

keywords = "Data fusion, Multisensor imagery, Multispectral data, Support vector machines (SVM), Synthetic aperture radar (SAR) data",

author = "Bj{\"o}rn Waske and Benediktsson, \{J{\'o}n Atli\}",

note = "Funding Information: B. Waske visited the Department of Electrical and Computer Engineering, University of Iceland, funded by the ENVILAND research project. He would like to thank G. Menz, M. Braun, and V. Heinzel from the University of Bonn. The software for the SVM classification was developed by A. Janz and S. Schiefer, Humboldt-Universit{\"a}t, Berlin, Germany. The authors would like to thank the European Space Agency for providing Envisat ASAR and ERS-2 data through a CAT 1 proposal (C1.3115). The SPOT image is provided through the European OASIS program (OASIS 58 - CE 6324). The data were acquired within the ENVILAND research project (FKZ 50EE0404), funded by the German Aerospace Center (DLR) and the Federal Ministry of Economics and Technology (BMWi). The authors would also like to thank the comments of the anonymous reviewers who helped us significantly improve this paper. Funding Information: Manuscript received December 10, 2006; revised March 19, 2007. This work was supported in part by the German Aerospace Center (DLR) and Federal Ministry of Economics and Technology (BMWi) under the project Enviland (FKZ 50EE0404), as well as an ESA Cat-1 proposals (C1P 3115) and European OASIS program (OASIS 58—CE 6324) and in part by the German Research Foundation (DFG) under the Research Training Group 722 (Information Techniques for Precision Crop Protection) at the University of Bonn.",

year = "2007",

month = dec,

doi = "10.1109/TGRS.2007.898446",

language = "English",

volume = "45",

pages = "3858--3866",

journal = "IEEE Transactions on Geoscience and Remote Sensing",

issn = "0196-2892",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "12",

}

TY - JOUR

T1 - Fusion of support vector machines for classification of multisensor data

AU - Waske, Björn

AU - Benediktsson, Jón Atli

N1 - Funding Information: B. Waske visited the Department of Electrical and Computer Engineering, University of Iceland, funded by the ENVILAND research project. He would like to thank G. Menz, M. Braun, and V. Heinzel from the University of Bonn. The software for the SVM classification was developed by A. Janz and S. Schiefer, Humboldt-Universität, Berlin, Germany. The authors would like to thank the European Space Agency for providing Envisat ASAR and ERS-2 data through a CAT 1 proposal (C1.3115). The SPOT image is provided through the European OASIS program (OASIS 58 - CE 6324). The data were acquired within the ENVILAND research project (FKZ 50EE0404), funded by the German Aerospace Center (DLR) and the Federal Ministry of Economics and Technology (BMWi). The authors would also like to thank the comments of the anonymous reviewers who helped us significantly improve this paper. Funding Information: Manuscript received December 10, 2006; revised March 19, 2007. This work was supported in part by the German Aerospace Center (DLR) and Federal Ministry of Economics and Technology (BMWi) under the project Enviland (FKZ 50EE0404), as well as an ESA Cat-1 proposals (C1P 3115) and European OASIS program (OASIS 58—CE 6324) and in part by the German Research Foundation (DFG) under the Research Training Group 722 (Information Techniques for Precision Crop Protection) at the University of Bonn.

PY - 2007/12

Y1 - 2007/12

N2 - The classification of multisensor data sets, consisting of multitemporal synthetic aperture radar data and optical imagery, is addressed. The concept is based on the decision fusion of different outputs. Each data source is treated separately and classified by a support vector machine (SVM). Instead of fusing the final classification outputs (i.e., land cover classes), the original outputs of each SVM discriminant function are used in the subsequent fusion process. This fusion is performed by another SVM, which is trained on the a priori outputs. In addition, two voting schemes are applied to create the final classification results. The results are compared with well-known parametric and nonparametric classifier methods, i.e., decision trees, the maximumlikelihood classifier, and classifier ensembles. The proposed SVM-based fusion approach outperforms all other approaches and significantly improves the results of a single SVM, which is trained on the whole multisensor data set.

AB - The classification of multisensor data sets, consisting of multitemporal synthetic aperture radar data and optical imagery, is addressed. The concept is based on the decision fusion of different outputs. Each data source is treated separately and classified by a support vector machine (SVM). Instead of fusing the final classification outputs (i.e., land cover classes), the original outputs of each SVM discriminant function are used in the subsequent fusion process. This fusion is performed by another SVM, which is trained on the a priori outputs. In addition, two voting schemes are applied to create the final classification results. The results are compared with well-known parametric and nonparametric classifier methods, i.e., decision trees, the maximumlikelihood classifier, and classifier ensembles. The proposed SVM-based fusion approach outperforms all other approaches and significantly improves the results of a single SVM, which is trained on the whole multisensor data set.

KW - Data fusion

KW - Multisensor imagery

KW - Multispectral data

KW - Support vector machines (SVM)

KW - Synthetic aperture radar (SAR) data

UR - https://www.scopus.com/pages/publications/36349007145

U2 - 10.1109/TGRS.2007.898446

DO - 10.1109/TGRS.2007.898446

M3 - Article

SN - 0196-2892

VL - 45

SP - 3858

EP - 3866

JO - IEEE Transactions on Geoscience and Remote Sensing

JF - IEEE Transactions on Geoscience and Remote Sensing

IS - 12

ER -

Fusion of support vector machines for classification of multisensor data

Abstract

Bibliographical note

Other keywords

Access to Document

Fingerprint

Cite this