Optimal band selection from hyperspectral data for 'Lantana camara' discrimination

Title
Optimal band selection from hyperspectral data for 'Lantana camara' discrimination
Publication Date
2012
Author(s)
Taylor, Subhashni
( author )
OrcID: https://orcid.org/0000-0002-1624-0901
Email: btaylo26@une.edu.au
UNE Id une-id:btaylo26
Kumar, Lalit
( author )
OrcID: https://orcid.org/0000-0002-9205-756X
Email: lkumar@une.edu.au
UNE Id une-id:lkumar
Reid, Nick
( author )
OrcID: https://orcid.org/0000-0002-4377-9734
Email: nrei3@une.edu.au
UNE Id une-id:nrei3
Lewis, Craig
Type of document
Journal Article
Language
en
Entity Type
Publication
Publisher
Taylor & Francis
Place of publication
United Kingdom
DOI
10.1080/01431161.2012.661093
UNE publication id
une:9832
Abstract
The primary objective of this research was to determine the optimal hyperspectral wavelengths based on spectroscopy data over the spectral range of 450-2500 nm for the detection of the invasive species 'Lantana camara' L. (lantana) from seven of its co-occurring species. A procedure based on statistical analysis of the reflectance and the first derivative reflectance (FDR) identified 86 and 18 bands, respectively, where lantana significantly differed from its co-occurring species. The effectiveness of the identified optimal bands was then evaluated using Hyperion imagery. The original Hyperion image with 155 bands gave an overall accuracy of 80% compared to 77% and 76% from the 86- and 18-band spectral subsets, respectively. A pairwise comparison of the three error matrices showed no significant difference in the accuracy achieved. The FDR analysis combined with the statistical analysis proved to be a useful procedure for data reduction by refining the discrimination to fewer optimal bands for lantana detection with no adverse impact on classification accuracy.
Link
Citation
International Journal of Remote Sensing, 33(17), p. 5418-5437
ISSN
1366-5901
0143-1161
Start page
5418
End page
5437

Files:

NameSizeformatDescriptionLink