Determining subcanopy Psidium cattleianum invasion in Hawaiian forests using imaging spectroscopy

Jomar M. Barbosa, Gregory P. Asner, Roberta E. Martin, Claire A. Baldeck, Flint Hughes, Tracy Johnson

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

High-resolution airborne imaging spectroscopy represents a promising avenue for mapping the spread of invasive tree species through native forests, but for this technology to be useful to forest managers there are two main technical challenges that must be addressed: (1) mapping a single focal species amongst a diverse array of other tree species; and (2) detecting early outbreaks of invasive plant species that are often hidden beneath the forest canopy. To address these challenges, we investigated the performance of two single-class classification frameworks-Biased Support Vector Machine (BSVM) and Mixture Tuned Matched Filtering (MTMF)-to estimate the degree of Psidium cattleianum incidence over a range of forest vertical strata (relative canopy density). We demonstrate that both BSVM and MTMF have the ability to detect relative canopy density of a single focal plant species in a vertically stratified forest, but they differ in the degree of user input required. Our results suggest BSVM as a promising method to disentangle spectrally-mixed classifications, as this approach generates decision values from a similarity function (kernel), which optimizes complex comparisons between classes using a dynamic machine learning process.

Original languageEnglish (US)
Article number33
JournalRemote Sensing
Volume8
Issue number1
DOIs
StatePublished - Jan 1 2016
Externally publishedYes

Fingerprint

spectroscopy
canopy
forest canopy
support vector machine
plant species

Keywords

  • Biased support vector machine
  • Carnegie airborne observatory
  • Invasive species
  • Mixture tuned matched filtering
  • Single-class classification
  • Strawberry guava

ASJC Scopus subject areas

  • Earth and Planetary Sciences(all)

Cite this

Determining subcanopy Psidium cattleianum invasion in Hawaiian forests using imaging spectroscopy. / Barbosa, Jomar M.; Asner, Gregory P.; Martin, Roberta E.; Baldeck, Claire A.; Hughes, Flint; Johnson, Tracy.

In: Remote Sensing, Vol. 8, No. 1, 33, 01.01.2016.

Research output: Contribution to journalArticle

Barbosa, Jomar M. ; Asner, Gregory P. ; Martin, Roberta E. ; Baldeck, Claire A. ; Hughes, Flint ; Johnson, Tracy. / Determining subcanopy Psidium cattleianum invasion in Hawaiian forests using imaging spectroscopy. In: Remote Sensing. 2016 ; Vol. 8, No. 1.
@article{0d5cdd77b63c4fa392dbcf44f1838ecc,
title = "Determining subcanopy Psidium cattleianum invasion in Hawaiian forests using imaging spectroscopy",
abstract = "High-resolution airborne imaging spectroscopy represents a promising avenue for mapping the spread of invasive tree species through native forests, but for this technology to be useful to forest managers there are two main technical challenges that must be addressed: (1) mapping a single focal species amongst a diverse array of other tree species; and (2) detecting early outbreaks of invasive plant species that are often hidden beneath the forest canopy. To address these challenges, we investigated the performance of two single-class classification frameworks-Biased Support Vector Machine (BSVM) and Mixture Tuned Matched Filtering (MTMF)-to estimate the degree of Psidium cattleianum incidence over a range of forest vertical strata (relative canopy density). We demonstrate that both BSVM and MTMF have the ability to detect relative canopy density of a single focal plant species in a vertically stratified forest, but they differ in the degree of user input required. Our results suggest BSVM as a promising method to disentangle spectrally-mixed classifications, as this approach generates decision values from a similarity function (kernel), which optimizes complex comparisons between classes using a dynamic machine learning process.",
keywords = "Biased support vector machine, Carnegie airborne observatory, Invasive species, Mixture tuned matched filtering, Single-class classification, Strawberry guava",
author = "Barbosa, {Jomar M.} and Asner, {Gregory P.} and Martin, {Roberta E.} and Baldeck, {Claire A.} and Flint Hughes and Tracy Johnson",
year = "2016",
month = "1",
day = "1",
doi = "10.3390/rs8010033",
language = "English (US)",
volume = "8",
journal = "Remote Sensing",
issn = "2072-4292",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "1",

}

TY - JOUR

T1 - Determining subcanopy Psidium cattleianum invasion in Hawaiian forests using imaging spectroscopy

AU - Barbosa, Jomar M.

AU - Asner, Gregory P.

AU - Martin, Roberta E.

AU - Baldeck, Claire A.

AU - Hughes, Flint

AU - Johnson, Tracy

PY - 2016/1/1

Y1 - 2016/1/1

N2 - High-resolution airborne imaging spectroscopy represents a promising avenue for mapping the spread of invasive tree species through native forests, but for this technology to be useful to forest managers there are two main technical challenges that must be addressed: (1) mapping a single focal species amongst a diverse array of other tree species; and (2) detecting early outbreaks of invasive plant species that are often hidden beneath the forest canopy. To address these challenges, we investigated the performance of two single-class classification frameworks-Biased Support Vector Machine (BSVM) and Mixture Tuned Matched Filtering (MTMF)-to estimate the degree of Psidium cattleianum incidence over a range of forest vertical strata (relative canopy density). We demonstrate that both BSVM and MTMF have the ability to detect relative canopy density of a single focal plant species in a vertically stratified forest, but they differ in the degree of user input required. Our results suggest BSVM as a promising method to disentangle spectrally-mixed classifications, as this approach generates decision values from a similarity function (kernel), which optimizes complex comparisons between classes using a dynamic machine learning process.

AB - High-resolution airborne imaging spectroscopy represents a promising avenue for mapping the spread of invasive tree species through native forests, but for this technology to be useful to forest managers there are two main technical challenges that must be addressed: (1) mapping a single focal species amongst a diverse array of other tree species; and (2) detecting early outbreaks of invasive plant species that are often hidden beneath the forest canopy. To address these challenges, we investigated the performance of two single-class classification frameworks-Biased Support Vector Machine (BSVM) and Mixture Tuned Matched Filtering (MTMF)-to estimate the degree of Psidium cattleianum incidence over a range of forest vertical strata (relative canopy density). We demonstrate that both BSVM and MTMF have the ability to detect relative canopy density of a single focal plant species in a vertically stratified forest, but they differ in the degree of user input required. Our results suggest BSVM as a promising method to disentangle spectrally-mixed classifications, as this approach generates decision values from a similarity function (kernel), which optimizes complex comparisons between classes using a dynamic machine learning process.

KW - Biased support vector machine

KW - Carnegie airborne observatory

KW - Invasive species

KW - Mixture tuned matched filtering

KW - Single-class classification

KW - Strawberry guava

UR - http://www.scopus.com/inward/record.url?scp=84957886510&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84957886510&partnerID=8YFLogxK

U2 - 10.3390/rs8010033

DO - 10.3390/rs8010033

M3 - Article

AN - SCOPUS:84957886510

VL - 8

JO - Remote Sensing

JF - Remote Sensing

SN - 2072-4292

IS - 1

M1 - 33

ER -