Douglas M. Flewelling, Ph.D.
James A. Ciarrocca, M.S.
This project presents an assessment tool for classifying building complexes using sitebased relationships as calculated from ArcGIS 9.2 using model builder and Python scripting. Anthropogenic features extracted from imagery often form the foundation of spatial databases. These data are in turn used to inform situational awareness for relief, law enforcement, and military agencies among many others. Buildings and the complexes they form are critical features within the landscape. The categorization of complexes requires an understanding of the relationships of the buildings within the site. In this study, building complexes in California were assessed for similarity to a prototypical California high school defined with a training set of known high schools and compared to a set of uncategorized sites. Eighty-eight percent of the high schools were correctly classified as being highly similar to the control data set.
Wilson, K. A. (2007). Building Complex and Site Categorization Using Similarity to a Prototypical Site (Master's thesis, University of Redlands). Retrieved from http://inspire.redlands.edu/gis_gradproj/93