Explicit signal to noise ratio in reproducing kernel Hilbert spaces

Luis Gomez-Chova; Allan Aasbjerg Nielsen; Gustavo Camps-Valls

doi:10.1109/IGARSS.2011.6049993

Explicit signal to noise ratio in reproducing kernel Hilbert spaces

Luis Gomez-Chova
, Allan Aasbjerg Nielsen
, Gustavo Camps-Valls

University of Valencia

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Abstract

This paper introduces a nonlinear feature extraction method based on kernels for remote sensing data analysis. The proposed approach is based on the minimum noise fraction (MNF) transform, which maximizes the signal variance while also minimizing the estimated noise variance. We here propose an alternative kernel MNF (KMNF) in which the noise is explicitly estimated in the reproducing kernel Hilbert space. This enables KMNF dealing with non-linear relations between the noise and the signal features jointly. Results show that the proposed KMNF provides the most noise-free features when confronted with PCA, MNF, KPCA, and the previous version of KMNF. Extracted features with the explicit KMNF also improve hyperspectral image classification.

Original language	English
Title of host publication	IEEE International Geoscience and Remote Sensing Symposium Proceedings (IGARSS)
Publisher	IEEE
Publication date	2011
Pages	3570-3573
ISBN (Print)	978-1-4577-1003-2
DOIs	https://doi.org/10.1109/IGARSS.2011.6049993
Publication status	Published - 2011
Event	2011 IEEE International Geoscience and Remote Sensing Symposium - Vancouver, Canada Duration: 24 Jul 2011 → 29 Jul 2011 https://ieeexplore.ieee.org/xpl/conhome/6034618/proceeding

Conference

Conference	2011 IEEE International Geoscience and Remote Sensing Symposium
Country/Territory	Canada
City	Vancouver
Period	24/07/2011 → 29/07/2011
Internet address	https://ieeexplore.ieee.org/xpl/conhome/6034618/proceeding

Access to Document

10.1109/IGARSS.2011.6049993

OpenUrl availability

Full text

Cite this

@inproceedings{c565de83db624ad89c89385435b85b72,

title = "Explicit signal to noise ratio in reproducing kernel Hilbert spaces",

abstract = "This paper introduces a nonlinear feature extraction method based on kernels for remote sensing data analysis. The proposed approach is based on the minimum noise fraction (MNF) transform, which maximizes the signal variance while also minimizing the estimated noise variance. We here propose an alternative kernel MNF (KMNF) in which the noise is explicitly estimated in the reproducing kernel Hilbert space. This enables KMNF dealing with non-linear relations between the noise and the signal features jointly. Results show that the proposed KMNF provides the most noise-free features when confronted with PCA, MNF, KPCA, and the previous version of KMNF. Extracted features with the explicit KMNF also improve hyperspectral image classification.",

author = "Luis Gomez-Chova and Nielsen, \{Allan Aasbjerg\} and Gustavo Camps-Valls",

year = "2011",

doi = "10.1109/IGARSS.2011.6049993",

language = "English",

isbn = "978-1-4577-1003-2",

pages = "3570--3573",

booktitle = "IEEE International Geoscience and Remote Sensing Symposium Proceedings (IGARSS)",

publisher = "IEEE",

address = "United States",

note = "2011 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2011 ; Conference date: 24-07-2011 Through 29-07-2011",

url = "https://ieeexplore.ieee.org/xpl/conhome/6034618/proceeding",

}

Gomez-Chova, L, Nielsen, AA & Camps-Valls, G 2011, Explicit signal to noise ratio in reproducing kernel Hilbert spaces. in IEEE International Geoscience and Remote Sensing Symposium Proceedings (IGARSS). IEEE, pp. 3570-3573, 2011 IEEE International Geoscience and Remote Sensing Symposium, Vancouver, British Columbia, Canada, 24/07/2011. https://doi.org/10.1109/IGARSS.2011.6049993

TY - GEN

T1 - Explicit signal to noise ratio in reproducing kernel Hilbert spaces

AU - Gomez-Chova, Luis

AU - Nielsen, Allan Aasbjerg

AU - Camps-Valls, Gustavo

PY - 2011

Y1 - 2011

N2 - This paper introduces a nonlinear feature extraction method based on kernels for remote sensing data analysis. The proposed approach is based on the minimum noise fraction (MNF) transform, which maximizes the signal variance while also minimizing the estimated noise variance. We here propose an alternative kernel MNF (KMNF) in which the noise is explicitly estimated in the reproducing kernel Hilbert space. This enables KMNF dealing with non-linear relations between the noise and the signal features jointly. Results show that the proposed KMNF provides the most noise-free features when confronted with PCA, MNF, KPCA, and the previous version of KMNF. Extracted features with the explicit KMNF also improve hyperspectral image classification.

AB - This paper introduces a nonlinear feature extraction method based on kernels for remote sensing data analysis. The proposed approach is based on the minimum noise fraction (MNF) transform, which maximizes the signal variance while also minimizing the estimated noise variance. We here propose an alternative kernel MNF (KMNF) in which the noise is explicitly estimated in the reproducing kernel Hilbert space. This enables KMNF dealing with non-linear relations between the noise and the signal features jointly. Results show that the proposed KMNF provides the most noise-free features when confronted with PCA, MNF, KPCA, and the previous version of KMNF. Extracted features with the explicit KMNF also improve hyperspectral image classification.

U2 - 10.1109/IGARSS.2011.6049993

DO - 10.1109/IGARSS.2011.6049993

M3 - Article in proceedings

SN - 978-1-4577-1003-2

SP - 3570

EP - 3573

BT - IEEE International Geoscience and Remote Sensing Symposium Proceedings (IGARSS)

PB - IEEE

T2 - 2011 IEEE International Geoscience and Remote Sensing Symposium

Y2 - 24 July 2011 through 29 July 2011

ER -

Explicit signal to noise ratio in reproducing kernel Hilbert spaces

Abstract

Conference

Access to Document

OpenUrl availability

Fingerprint

Cite this