DocumentCode :
483945
Title :
Mapping Riparian and Wetland Weeds with High Resolution Satellite Imagery
Author :
Sriharan, Shobha ; Everitt, James H. ; Yang, Chenghai ; Fletcher, Reginald S.
Author_Institution :
Virginia State Univ., Petersburg, VA
Volume :
1
fYear :
2008
fDate :
7-11 July 2008
Abstract :
QuickBird multi-spectral satellite imagery was evaluated for detecting weeds in waterways and wetlands in Texas. Plant species studied included giant reed (Arundo donax L.), waterhyacinth [Eichhornia crassipes (Mart.) Solms], and giant salvinia (Salvinia molesta Mitchell). Accuracy assessment results performed on the unsupervised classified maps for each species were: giant reed, 94.4% producer´s accuracy and 100% user´s accuracy; waterhyacinth, 100% producer´s accuracy and 80% user´s accuracy; and giant salvinia, 93.9% producer´s accuracy and 92% user´s accuracy. These results indicate that QuickBird imagery can be used successfully to map infestations of these three weeds.
Keywords :
image classification; image processing; vegetation mapping; Arundo donax L; Eichhornia crassipes; QuickBird imagery; Salyinia molesta Mitchell; Texas; USA; Water hyacinth; giant reed; giant salvinia; high resolution satellite imagery; image classification; image processing; multi-spectral satellite imagery; plant species; riparian weed mapping; weed detection; wetland weed mapping; Brushes; Government; Image color analysis; Image resolution; Protection; Remote sensing; Satellites; Soil; Spatial resolution; US Department of Agriculture;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2008. IGARSS 2008. IEEE International
Conference_Location :
Boston, MA
Print_ISBN :
978-1-4244-2807-6
Electronic_ISBN :
978-1-4244-2808-3
Type :
conf
DOI :
10.1109/IGARSS.2008.4778899
Filename :
4778899
Link To Document :
بازگشت