Abstract
Mapping their location and extent is a critical step in noxious weed management. One of the most common methods of mapping noxious weeds is to walk the perimeter of each patch with a handheld GPS receiver. This is the method used at Camp Williams, a National Guard Bureau training facility in Utah where this study was conducted. It was proposed that a low-cost Unmanned Aerial Vehicle (UAV) that made use of a hobbyist remote control airplane equipped with a Global Positioning System (GPS) receiver and digital camera could be used along with automated post-processing techniques to reduce the cost of weed mapping compared to the on foot method. Two noxious weeds were studied: musk thistle (Carduus nutans) and dalmation toadflax (Linaria dalmatica). The musk thistle was visually identifiable in the imagery but the dalmation toadflax was confused with yellow sweet clover (Melilotus officinalis). It was found that after the automated post-processing the photos were not positioned well enough to produce a consistent and accurate weed perimeter. A supervised classification was attempted with imagery of the musk thistle, however, the accuracy of the classification was too low to be able to identify the weed perimeter from the classification. To achieve accurate results the photos had to be registered to a base image and the perimeter of each patch hand digitized. The time it took to do so increased the costs well above the on foot method. A number of improvements to the UAV could make the image registration step unnecessary. There are other applications for which this low cost UAV could be used.
Degree
MS
College and Department
Family, Home, and Social Sciences; Geography
Rights
http://lib.byu.edu/about/copyright/
BYU ScholarsArchive Citation
Jones, Brandon Tyler, "An Evaluation of a Low-Cost UAV Approach to Noxious Weed Mapping" (2007). Theses and Dissertations. 1220.
https://scholarsarchive.byu.edu/etd/1220
Date Submitted
2007-11-20
Document Type
Thesis
Handle
http://hdl.lib.byu.edu/1877/etd2146
Keywords
UAV, Noxious Weeds, Weed Mapping, Land Management, Remote Sensing, GIS, Weed Management
Language
English