Abstract:
The Oregon Department of Fish and Wildlife's (ODFW) Ecological Analysis
Center (EAC) is in the process of creating, from Landsat Thematic Mapper (TM)
imagery, a vegetation map of Oregon that will meet the latest standards set by the
National Gap Analysis Program. Since field verification is often expensive and by nature
intensive, ODFW wanted to determine the feasibility of using airborne videography to
help classify and validate their Oregon vegetation map. In 1993, ODFW sampled
approximately 4% of Oregon with airborne videography by flying north-south transects at
30-km intervals to be used for these purposes.
This research was designed to examine how and to what extent airborne
videography can be used for assessing the accuracy of classified satellite imagery in
vegetation mapping. An accuracy assessment strategy incorporating the ODFW airborne
videography was developed and tested on a pilot study area consisting of the Luckiamute
and Rickreall watersheds in western Oregon.
Airborne videography was found to have more potential as a classification aid
than as an accuracy assessment tool. Its limited usefulness in accuracy assessment results
primarily from the necessity to field verify any interpretation made of the videography before it can be successfully incorporated into an accuracy. assessment methodology.
Additionally, the difficulty of obtaining a sufficient sample for all vegetation classes and
the relatively poor spatial and spectral resolution of current airborne video systems
impede its use in accuracy assessment. A field verification process combining global
positioning system (GPS) and geographic information system (GIS) technologies with a
laptop computer is outlined as a more efficient and accurate alternative to using the
ODFW airborne videography for accuracy assessment of the Oregon vegetation map.
