Project Title:

Auto-Vectorization of Areal- and Linear-Raster Image Features

07.03-1813
910778
Auto-Vectorization of Areal- and Linear-Raster Image Features
Delta Data Systems, Inc.
131 Third Street
Picayune
MS
39466
Andrew A.
Rost
601-799-1813
SSC
NAS13-481
122

07.03-1813
910778

Abstract:

Auto-Vectorization of Areal- and Linear-Raster Image Features
Remotely sensed images contain areal and linear features that may complement, but
are generally not available to, geographic information systems (GIS). Auto-vectorization
transforms the tesselated data structure of raster images into the topological data
structure of vector-based GISs. This technology can be applied to any raster image
including scanned maps. Auto-vectorization combined with auto-attribution and expert
system technology will produce an autonomous system for the conversion of maps and
remotely sensed images to "GIS-ready" objects. The primary innovations of this project
include the integration of remotely sensed data with mainstream GIS packages, the
reduction of cost, time, and labor associated with GIS development, and the automatic
extraction of linear earth features from remotely sensed images.
Applications would exist in expert systems to populate GIS data bases; mapping systems
based on computer visualization and/or analysis (i.e., coastline for fault mapping);
and resource studies such as tree crop inventory and oil-bearing rock porosity.
auto-vectorization, topology, tessellation, theme-island, auto-attribution, visualization,
data integration