Single-Species Detection with Airborne Imaging Spectroscopy Data: A Comparison of Support Vector Techniques

Claire A. Baldeck, Gregory P. Asner

Research output: Contribution to journalArticle

27 Citations (Scopus)

Abstract

Progress in mapping plant species remotely with imaging spectroscopy data is limited by the traditional classification framework, which carries the requirement of exhaustively defining all classes (species) encountered in a landscape. As the research objective may be to map only one or a few species of interest, we need to explore alternative classification methods that may be used to more efficiently detect a single species. We compared the performance of three support vector machine (SVM) methods designed for single-class detection - binary (one-against-all) SVM, one-class SVM, and biased SVM - in detecting five focal tree and shrub species using data collected by the Carnegie Airborne Observatory over an African savanna. Prior to this comparison, we investigated the effects of training data amount and balance on binary SVM and evaluated alternative methods for tuning one-class and biased SVMs. A key finding was that biased SVM was generally best parameterized by crown-level cross validation paired with the tuning criterion proposed by Lee and Liu [1]. Among the different single-class methods, binary SVM showed the best overall performance (average F-scores 0.43-0.78 among species), whereas one-class SVM showed very poor performance (F-scores 0.09-0.46). However, biased SVM produced results similar to those obtained with binary SVM (F-scores 0.40-0.72), despite using labeled training data from only the focal class. Our results indicate that both binary and biased SVMs can work well for remote single-species detection, while both methods, particularly biased SVM, greatly reduce the amount of training data required compared with traditional multispecies classification.

Original languageEnglish (US)
Article number6891145
Pages (from-to)2501-2512
Number of pages12
JournalIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Volume8
Issue number6
DOIs
StatePublished - Jun 1 2015
Externally publishedYes

Fingerprint

Support vector machines
spectroscopy
Spectroscopy
Imaging techniques
detection
comparison
support vector machine
Tuning
Observatories
savanna
shrub
observatory
method

Keywords

  • Biased SVM
  • hyperspectral remote sensing
  • one-class SVM
  • remote species identification
  • single-class classification
  • SVM

ASJC Scopus subject areas

  • Computers in Earth Sciences
  • Atmospheric Science

Cite this

@article{8a1d0de0ccdb473a970f8de458dfd00f,
title = "Single-Species Detection with Airborne Imaging Spectroscopy Data: A Comparison of Support Vector Techniques",
abstract = "Progress in mapping plant species remotely with imaging spectroscopy data is limited by the traditional classification framework, which carries the requirement of exhaustively defining all classes (species) encountered in a landscape. As the research objective may be to map only one or a few species of interest, we need to explore alternative classification methods that may be used to more efficiently detect a single species. We compared the performance of three support vector machine (SVM) methods designed for single-class detection - binary (one-against-all) SVM, one-class SVM, and biased SVM - in detecting five focal tree and shrub species using data collected by the Carnegie Airborne Observatory over an African savanna. Prior to this comparison, we investigated the effects of training data amount and balance on binary SVM and evaluated alternative methods for tuning one-class and biased SVMs. A key finding was that biased SVM was generally best parameterized by crown-level cross validation paired with the tuning criterion proposed by Lee and Liu [1]. Among the different single-class methods, binary SVM showed the best overall performance (average F-scores 0.43-0.78 among species), whereas one-class SVM showed very poor performance (F-scores 0.09-0.46). However, biased SVM produced results similar to those obtained with binary SVM (F-scores 0.40-0.72), despite using labeled training data from only the focal class. Our results indicate that both binary and biased SVMs can work well for remote single-species detection, while both methods, particularly biased SVM, greatly reduce the amount of training data required compared with traditional multispecies classification.",
keywords = "Biased SVM, hyperspectral remote sensing, one-class SVM, remote species identification, single-class classification, SVM",
author = "Baldeck, {Claire A.} and Asner, {Gregory P.}",
year = "2015",
month = "6",
day = "1",
doi = "10.1109/JSTARS.2014.2346475",
language = "English (US)",
volume = "8",
pages = "2501--2512",
journal = "IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing",
issn = "1939-1404",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "6",

}

TY - JOUR

T1 - Single-Species Detection with Airborne Imaging Spectroscopy Data

T2 - A Comparison of Support Vector Techniques

AU - Baldeck, Claire A.

AU - Asner, Gregory P.

PY - 2015/6/1

Y1 - 2015/6/1

N2 - Progress in mapping plant species remotely with imaging spectroscopy data is limited by the traditional classification framework, which carries the requirement of exhaustively defining all classes (species) encountered in a landscape. As the research objective may be to map only one or a few species of interest, we need to explore alternative classification methods that may be used to more efficiently detect a single species. We compared the performance of three support vector machine (SVM) methods designed for single-class detection - binary (one-against-all) SVM, one-class SVM, and biased SVM - in detecting five focal tree and shrub species using data collected by the Carnegie Airborne Observatory over an African savanna. Prior to this comparison, we investigated the effects of training data amount and balance on binary SVM and evaluated alternative methods for tuning one-class and biased SVMs. A key finding was that biased SVM was generally best parameterized by crown-level cross validation paired with the tuning criterion proposed by Lee and Liu [1]. Among the different single-class methods, binary SVM showed the best overall performance (average F-scores 0.43-0.78 among species), whereas one-class SVM showed very poor performance (F-scores 0.09-0.46). However, biased SVM produced results similar to those obtained with binary SVM (F-scores 0.40-0.72), despite using labeled training data from only the focal class. Our results indicate that both binary and biased SVMs can work well for remote single-species detection, while both methods, particularly biased SVM, greatly reduce the amount of training data required compared with traditional multispecies classification.

AB - Progress in mapping plant species remotely with imaging spectroscopy data is limited by the traditional classification framework, which carries the requirement of exhaustively defining all classes (species) encountered in a landscape. As the research objective may be to map only one or a few species of interest, we need to explore alternative classification methods that may be used to more efficiently detect a single species. We compared the performance of three support vector machine (SVM) methods designed for single-class detection - binary (one-against-all) SVM, one-class SVM, and biased SVM - in detecting five focal tree and shrub species using data collected by the Carnegie Airborne Observatory over an African savanna. Prior to this comparison, we investigated the effects of training data amount and balance on binary SVM and evaluated alternative methods for tuning one-class and biased SVMs. A key finding was that biased SVM was generally best parameterized by crown-level cross validation paired with the tuning criterion proposed by Lee and Liu [1]. Among the different single-class methods, binary SVM showed the best overall performance (average F-scores 0.43-0.78 among species), whereas one-class SVM showed very poor performance (F-scores 0.09-0.46). However, biased SVM produced results similar to those obtained with binary SVM (F-scores 0.40-0.72), despite using labeled training data from only the focal class. Our results indicate that both binary and biased SVMs can work well for remote single-species detection, while both methods, particularly biased SVM, greatly reduce the amount of training data required compared with traditional multispecies classification.

KW - Biased SVM

KW - hyperspectral remote sensing

KW - one-class SVM

KW - remote species identification

KW - single-class classification

KW - SVM

UR - http://www.scopus.com/inward/record.url?scp=85027952483&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85027952483&partnerID=8YFLogxK

U2 - 10.1109/JSTARS.2014.2346475

DO - 10.1109/JSTARS.2014.2346475

M3 - Article

AN - SCOPUS:85027952483

VL - 8

SP - 2501

EP - 2512

JO - 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

M1 - 6891145

ER -