Developing a Quantitative Accuracy Assessment Application for ArcView GIS
Track: Remote Sensing Imagery
Author(s): Rick Jones, Aaron Heriford
Since GIS is simply a model of reality, measuring data quality is of premier importance. Over the past several years, Pacific Meridian Resources has proactively defined the industry standard for quantitative accuracy assessment. To facilitate the application of these techniques for developing large land cover databases, an ArcView GIS extension has been created to aid the GIS specialist in evaluating the quality of remotely sensed classified data. This paper reviews the achievements and issues related to development of the custom ArcView application for accuracy assessment. The custom application provides a three-tier solution to an accuracy assessment project, namely (1) random sampling of base data, (2) assessment of random samples, and (3) error matrix generation. Tools are provided in the application to guide the user through each step in the accuracy assessment process. This paper will demonstrate the use of the ArcView Accuracy Assessment extension to conduct an accuracy assessment of land cover classes from classified imagery, demonstrating user interaction with custom dialogs and wizards.
Pacific Meridian Resources
421 S.W. 6th Avenue, Suite 850
Portland, OR 97204