Final report




Дата канвертавання15.04.2016
Памер12.46 Kb.


Using a Remotely Piloted Vehicle to detect Squarrose Knapweed (Centaurea virgata Lam. Ssp. squarrosa Gugl.) at

Rangeland Sites in Utah
Perry J. Hardin, Mark. W. Jackson, Val J. Anderson, and Robert Johnson
FINAL REPORT

Research Grant GC253-04-Z1138


December 20, 2005
Perry J. Hardin (Primary contact)

Assoc. Professor of Geography

Brigham Young University

690 SWKT Phone: 801.422.3851

Provo, Utah 84602 Email: perry_hardin@byu.edu
Mark W. Jackson

Asst. Professor of Geography

Brigham Young University

622 SWKT Phone: 801.422.9753

Provo, Utah 84602 Email: mj@byu.edu
Val J. Anderson

Professor of Integrative Biology

Brigham Young University

493 WIDB Phone: 801.422.3527

Provo, Utah 84602 Email: val_anderson@byu.edu
Robert Johnson

Department of Integrative Biology

Brigham Young University

401 WIDB Phone: 801.422.3311

Provo, Utah 84602 Email: robert_johnson@byu.edu

Table of Contents
Executive Summary iii
Introduction 1
Data and Methods 3
Results and Discussion 28
Concluding Comments and Recommendations 41
Bibliography 45
Compact Disk Contents 47

Using a Remotely Piloted Vehicle to detect Squarrose Knapweed (Centaurea virgata Lam. Ssp. squarrosa Gugl.) at Rangeland Sites in Utah
EXECUTIVE SUMMARY
Research Grant GC253-04-Z1138
Perry J. Hardin, Mark. W. Jackson, Val J. Anderson, and Robert Johnson
December 20, 2005
The squarrose knapweed (Centaurea virgata Lam. Ssp. squarrosa Gugl.) shrub is native to southwest Asia and the Middle East. It was probably introduced to the U.S. in the early 20th century. It is considered an invasive weedy species, and is noxious in Utah.

This research began with the working hypothesis that imagery acquired from inexpensive RPVs could be used to detect and map patches of squarrose knapweed along invasion corridors such as roads, paths, and riparian areas. The three research objectives were: 1) to determine whether digital imagery acquired from inexpensive RPVs contain sufficient information to detect squarrose knapweed, 2) to determine the factors which maximize the chances of detecting and mapping squarrose knapweed from digital RPV imagery, and 3) to compare the effectiveness of an RPV squarrose knapweed survey against a traditional windshield survey.

This research shows that a significant proportion of squarrose knapweed patches can be accurately detected on photography acquired from low flying RPVs. Four rangeland study sites were examined. Under ideal conditions at all four sites, approximately 50% of the squarrose knapweed patches were detectable on RPV photography acquired from an altitude of 100 meters above ground level. False alarms were negligible. In the two remote desert study sites, RPV-based squarrose knapweed detection rates were superior to detection rates obtained through a traditional windshield survey. This was primarily due to the vertical aerial perspective offered by the RPV that provided visibility in areas obscured by terrain or vegetation to the terrestrial surveyor. As expected, in the two study sites where the knapweed was immediately adjacent to a paved highway (and only a few meters away from the windshield survey observer), the RPV detection rate was lower than the rate obtained through the windshield survey.

The squarrose knapweed detection rates varied from month to month. Because its color resembled that of surrounding vegetation, there was marginal success in mapping the squarrose knapweed with the RPV in the spring. In summer, detection capability was significant. The increase in detection rates from spring to summer were due to: 1) a progressive increase in knapweed shrub size/branching density, and 2) a relative increase in the target-background contrast as the grasses browned and the knapweed remained green. The squarrose knapweed could be detected most frequently in midsummer (July), but some detection capability existed in the early summer long before the knapweed flowered or seedheads matured. This shows great promise for locating small patches of mature squarrose knapweed with the RPV before infestations can spread.

While the ability to map squarrose knapweed from the RPV has been demonstrated, the cost effectiveness of the RPV approach depends on the precise mission defined for the RPV. If the future purpose of using the RPV was merely to detect any knapweed on a photograph without regard for careful counts or precise mapping, the RPV approach could probably be cost effective. If any knapweed was detected while quickly scanning the RPV photographs, the global position coordinates of the photograph could then be recorded for later use by the weed eradication field crew.

Although the system can be used to detect knapweed patches and might be cost effective, we do not consider the digital RPV aircraft used in this research to be a practical solution to detecting invasive weeds in remote areas. The most significant problem is that the RPV flies too fast for a novice pilot to control during landing. Ground handling of the RPV during taxi and takeoff is also difficult. It is also likewise problematic for a novice pilot to maintain the aircraft at a constant altitude while the pictures are being acquired. The research team is currently designing an RPV system with improvements that should ameliorate these problems, but more development and testing will be required to increase the feasibility of its use by range technicians who must fly it on a regular basis.






База данных защищена авторским правом ©shkola.of.by 2016
звярнуцца да адміністрацыі

    Галоўная старонка