Fast implementation of singular spectrum analysis for effective feature extraction in hyperspectral imaging

Jaime Zabalza, Jinchang Ren, Zheng Wang, Huimin Zhao, Jun Wang, Stephen Marshall

Research output: Contribution to journalArticle

15 Citations (Scopus)

Abstract

As a recent approach for time series analysis, singular spectrum analysis (SSA) has been successfully applied for feature extraction in hyperspectral imaging (HSI), leading to increased accuracy in pixel-based classification tasks. However, one of the main drawbacks of conventional SSA in HSI is the extremely high computational complexity, where each pixel requires individual and complete singular value decomposition (SVD) analyses. To address this issue, a fast implementation of SSA (F-SSA) is proposed for efficient feature extraction in HSI. Rather than applying pixel-based SVD as conventional SSA does, the fast implementation only needs one SVD applied to a representative pixel, i.e., either the median or the mean spectral vector of the HSI hypercube. The result of SVD is employed as a unique transform matrix for all the pixels within the hypercube. As demonstrated in experiments using two well-known publicly available data sets, almost identical results are produced by the fast implementation in terms of accuracy of data classification, using the support vector machine (SVM) classifier. However, the overall computational complexity has been significantly reduced.
LanguageEnglish
Pages2845-2853
Number of pages9
JournalIEEE Journal of Selected Topics in Earth Observation and Remote Sensing
Volume8
Issue number6
Early online date19 Dec 2014
DOIs
Publication statusPublished - Jun 2015

Fingerprint

Spectrum analysis
Feature extraction
Singular value decomposition
pixel
Pixels
decomposition
Computational complexity
Time series analysis
time series analysis
Support vector machines
Classifiers
transform
Hyperspectral imaging
analysis
matrix
experiment
Experiments

Keywords

  • data classification
  • fast singular spectrum analysis
  • feature extraction
  • hyperspectral imaging
  • support vector machine

Cite this

@article{17c8bed5eec94b9cbb8fa0328dfda8d2,
title = "Fast implementation of singular spectrum analysis for effective feature extraction in hyperspectral imaging",
abstract = "As a recent approach for time series analysis, singular spectrum analysis (SSA) has been successfully applied for feature extraction in hyperspectral imaging (HSI), leading to increased accuracy in pixel-based classification tasks. However, one of the main drawbacks of conventional SSA in HSI is the extremely high computational complexity, where each pixel requires individual and complete singular value decomposition (SVD) analyses. To address this issue, a fast implementation of SSA (F-SSA) is proposed for efficient feature extraction in HSI. Rather than applying pixel-based SVD as conventional SSA does, the fast implementation only needs one SVD applied to a representative pixel, i.e., either the median or the mean spectral vector of the HSI hypercube. The result of SVD is employed as a unique transform matrix for all the pixels within the hypercube. As demonstrated in experiments using two well-known publicly available data sets, almost identical results are produced by the fast implementation in terms of accuracy of data classification, using the support vector machine (SVM) classifier. However, the overall computational complexity has been significantly reduced.",
keywords = "data classification, fast singular spectrum analysis, feature extraction, hyperspectral imaging, support vector machine",
author = "Jaime Zabalza and Jinchang Ren and Zheng Wang and Huimin Zhao and Jun Wang and Stephen Marshall",
note = "(c) 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works.",
year = "2015",
month = "6",
doi = "10.1109/JSTARS.2014.2375932",
language = "English",
volume = "8",
pages = "2845--2853",
journal = "IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing",
issn = "1939-1404",
publisher = "IEEE",
number = "6",

}

TY - JOUR

T1 - Fast implementation of singular spectrum analysis for effective feature extraction in hyperspectral imaging

AU - Zabalza, Jaime

AU - Ren, Jinchang

AU - Wang, Zheng

AU - Zhao, Huimin

AU - Wang, Jun

AU - Marshall, Stephen

N1 - (c) 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works.

PY - 2015/6

Y1 - 2015/6

N2 - As a recent approach for time series analysis, singular spectrum analysis (SSA) has been successfully applied for feature extraction in hyperspectral imaging (HSI), leading to increased accuracy in pixel-based classification tasks. However, one of the main drawbacks of conventional SSA in HSI is the extremely high computational complexity, where each pixel requires individual and complete singular value decomposition (SVD) analyses. To address this issue, a fast implementation of SSA (F-SSA) is proposed for efficient feature extraction in HSI. Rather than applying pixel-based SVD as conventional SSA does, the fast implementation only needs one SVD applied to a representative pixel, i.e., either the median or the mean spectral vector of the HSI hypercube. The result of SVD is employed as a unique transform matrix for all the pixels within the hypercube. As demonstrated in experiments using two well-known publicly available data sets, almost identical results are produced by the fast implementation in terms of accuracy of data classification, using the support vector machine (SVM) classifier. However, the overall computational complexity has been significantly reduced.

AB - As a recent approach for time series analysis, singular spectrum analysis (SSA) has been successfully applied for feature extraction in hyperspectral imaging (HSI), leading to increased accuracy in pixel-based classification tasks. However, one of the main drawbacks of conventional SSA in HSI is the extremely high computational complexity, where each pixel requires individual and complete singular value decomposition (SVD) analyses. To address this issue, a fast implementation of SSA (F-SSA) is proposed for efficient feature extraction in HSI. Rather than applying pixel-based SVD as conventional SSA does, the fast implementation only needs one SVD applied to a representative pixel, i.e., either the median or the mean spectral vector of the HSI hypercube. The result of SVD is employed as a unique transform matrix for all the pixels within the hypercube. As demonstrated in experiments using two well-known publicly available data sets, almost identical results are produced by the fast implementation in terms of accuracy of data classification, using the support vector machine (SVM) classifier. However, the overall computational complexity has been significantly reduced.

KW - data classification

KW - fast singular spectrum analysis

KW - feature extraction

KW - hyperspectral imaging

KW - support vector machine

UR - http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=4609443

U2 - 10.1109/JSTARS.2014.2375932

DO - 10.1109/JSTARS.2014.2375932

M3 - Article

VL - 8

SP - 2845

EP - 2853

JO - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

T2 - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

JF - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

SN - 1939-1404

IS - 6

ER -