Sparse principal component analysis in hyperspectral change detection

Allan Aasbjerg Nielsen; Rasmus Larsen; Jacob Schack Vestergaard

doi:10.1117/12.897434

Sparse principal component analysis in hyperspectral change detection

Allan Aasbjerg Nielsen, Rasmus Larsen, Jacob Schack Vestergaard

Research output: Contribution to journal › Conference article › Research

3 Downloads (Pure)

Abstract

This contribution deals with change detection by means of sparse principal component analysis (PCA) of simple differences of calibrated, bi-temporal HyMap data. Results show that if we retain only 15 nonzero loadings (out of 126) in the sparse PCA the resulting change scores appear visually very similar although the loadings are very different from their usual non-sparse counterparts. The choice of three wavelength regions as being most important for change detection demonstrates the feature selection capability of sparse PCA.

Original language	English
Journal	Proceedings of SPIE - The International Society for Optical Engineering
Volume	8180
Pages (from-to)	81800S
ISSN	0277-786X
DOIs	https://doi.org/10.1117/12.897434
Publication status	Published - 2011
Event	Image and Signal Processing for Remote Sensing XVII - Prague, Czech Republic Duration: 19 Sept 2011 → 22 Sept 2011

Conference

Conference	Image and Signal Processing for Remote Sensing XVII
Country/Territory	Czech Republic
City	Prague
Period	19/09/2011 → 22/09/2011

Keywords

Feature selection
HyMap
Airborne remote sensing

Access to Document

10.1117/12.897434

OpenUrl availability

Full text

Cite this

@inproceedings{5d5f19dc41b743f19fc2525d96db9694,

title = "Sparse principal component analysis in hyperspectral change detection",

abstract = "This contribution deals with change detection by means of sparse principal component analysis (PCA) of simple differences of calibrated, bi-temporal HyMap data. Results show that if we retain only 15 nonzero loadings (out of 126) in the sparse PCA the resulting change scores appear visually very similar although the loadings are very different from their usual non-sparse counterparts. The choice of three wavelength regions as being most important for change detection demonstrates the feature selection capability of sparse PCA.",

keywords = "Feature selection, HyMap, Airborne remote sensing",

author = "Nielsen, {Allan Aasbjerg} and Rasmus Larsen and Vestergaard, {Jacob Schack}",

year = "2011",

doi = "10.1117/12.897434",

language = "English",

volume = "8180",

pages = "81800S",

journal = "Proceedings of SPIE - The International Society for Optical Engineering",

issn = "0277-786X",

publisher = "Society of Photo-Optical Instrumentation Engineers",

note = "Image and Signal Processing for Remote Sensing XVII ; Conference date: 19-09-2011 Through 22-09-2011",

}

TY - GEN

T1 - Sparse principal component analysis in hyperspectral change detection

AU - Nielsen, Allan Aasbjerg

AU - Larsen, Rasmus

AU - Vestergaard, Jacob Schack

PY - 2011

Y1 - 2011

N2 - This contribution deals with change detection by means of sparse principal component analysis (PCA) of simple differences of calibrated, bi-temporal HyMap data. Results show that if we retain only 15 nonzero loadings (out of 126) in the sparse PCA the resulting change scores appear visually very similar although the loadings are very different from their usual non-sparse counterparts. The choice of three wavelength regions as being most important for change detection demonstrates the feature selection capability of sparse PCA.

AB - This contribution deals with change detection by means of sparse principal component analysis (PCA) of simple differences of calibrated, bi-temporal HyMap data. Results show that if we retain only 15 nonzero loadings (out of 126) in the sparse PCA the resulting change scores appear visually very similar although the loadings are very different from their usual non-sparse counterparts. The choice of three wavelength regions as being most important for change detection demonstrates the feature selection capability of sparse PCA.

KW - Feature selection

KW - HyMap

KW - Airborne remote sensing

U2 - 10.1117/12.897434

DO - 10.1117/12.897434

M3 - Conference article

SN - 0277-786X

VL - 8180

SP - 81800S

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

T2 - Image and Signal Processing for Remote Sensing XVII

Y2 - 19 September 2011 through 22 September 2011

ER -

Sparse principal component analysis in hyperspectral change detection

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this