Skip to main content
ARS Home » Research » Publications at this Location » Publication #271577
Title: Applying spectral unmixing and support vector machine to airborne hyperspectral imagery for detecting giant reed

Author
item Yang, Chenghai
item Goolsby, John
item EVERITT, JAMES - Retired ARS Employee
item DU, QIAN - Mississippi State University

Submitted to: IEEE IGARSS Annual Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 8/25/2011
Publication Date: 9/15/2011
Citation: Yang, C., Goolsby, J., Everitt, J.H., Du, Q. 2011. Applying spectral unmixing and support vector machine to airborne hyperspectral imagery for detecting giant reed. IEEE IGARSS Annual Proceedings. 2011 CDROM.

Interpretive Summary: Giant reed is an invasive weed that presents a severe threat to agroecosystems and riparian areas throughout the southern United States and northern Mexico with the densest stands growing along the Rio Grande in Texas and the coastal rivers of southern California. Accurate information on the spatial distribution and infested areas of giant reed is essential for effective management of this invasive weed. This study evaluated three image classification techniques to identify giant reed from airborne hyperspectral imagery taken from a site along the US-Mexican portion of the Rio Grande in 2009 and 2010. Accuracy assessment showed that these techniques applied to airborne hyperspectral imagery can be used to effectively distinguish giant reed from associated plant species and to monitor the progression of this invasive weed.

Technical Abstract: This study evaluated linear spectral unmixing (LSU), mixture tuned matched filtering (MTMF) and support vector machine (SVM) techniques for detecting and mapping giant reed (Arundo donax L.), an invasive weed that presents a severe threat to agroecosystems and riparian areas throughout the southern United States and northern Mexico. Airborne hyperspectral imagery with 102 usable bands covering a spectral range of 475-845 nm was collected from a giant reed-infested site along the US-Mexican portion of the Rio Grande in 2009 and 2010. The imagery was transformed with minimum noise fraction (MFN) to reduce the spectral dimensionality and noise. The three classification techniques (LSU, MTMF and SVM) were applied to the transformed MNF imagery based 11 endmember spectra extracted from the images for each of the two years. Accuracy assessment and kappa analysis were performed to compare the differences in classification accuracies among the three classification methods. Results showed that SVM and MTMF performed better than LSU, with SVM being the best classifier in both years. The results from this study indicate that hyperspectral imagery in conjunction with image classification techniques is useful for distinguishing giant reed from associated plant species and for monitoring the progression of this invasive weed.