Share Email Print

Journal of Applied Remote Sensing

Object-based classification of semi-arid wetlands
Author(s): Meghan Halabisky; L. Monika Moskal; Sonia A. Hall
Format Member Price Non-Member Price
PDF $20.00 $25.00

Paper Abstract

Wetlands are valuable ecosystems that benefit society. However, throughout history wetlands have been converted to other land uses. For this reason, timely wetland maps are necessary for developing strategies to protect wetland habitat. The goal of this research was to develop a time-efficient, automated, low-cost method to map wetlands in a semi-arid landscape that could be scaled up for use at a county or state level, and could lay the groundwork for expanding to forested areas. Therefore, it was critical that the research project contain two components: accurate automated feature extraction and the use of low-cost imagery. For that reason, we tested the effectiveness of geographic object-based image analysis (GEOBIA) to delineate and classify wetlands using freely available true color aerial photographs provided through the National Agriculture Inventory Program. The GEOBIA method produced an overall accuracy of 89% (khat = 0.81), despite the absence of infrared spectral data. GEOBIA provides the automation that can save significant resources when scaled up while still providing sufficient spatial resolution and accuracy to be useful to state and local resource managers and policymakers.

Paper Details

Date Published: 1 January 2011
PDF: 14 pages
J. Appl. Remote Sens. 5(1) 053511 doi: 10.1117/1.3563569
Published in: Journal of Applied Remote Sensing Volume 5, Issue 1
Show Author Affiliations
Meghan Halabisky, Univ. of Washington (United States)
L. Monika Moskal, Univ. of Washington (United States)
Sonia A. Hall, The Nature Conservancy (United States)

© SPIE. Terms of Use
Back to Top