News
Release
Geosemble
Awarded NSF Grant for GeoPrism Satellite-based Direct Marketing
As part of the program, Geosemble Technologies, in
collaboration with the University of Southern California, will build upon its
work in map fusion and imagery analysis to automatically extract image data from
satellite imagery, link it to property address data, then cross reference with
open source online data to extract highly targeted marketing leads of interest
to a range of manufacturers and service providers.
Supported by the National Science Foundation, Geosemble
Technologies will leverage three key technologies in carrying out the GeoPrism
program. First, the company has
developed machine-learning techniques for Automatic Feature Extraction of
satellite imagery, allowing GeoPrism to “understand” and classify what it is
seeing. Second, the company has
patented its Digital Rubber Sheeting technique, creating the capability for
GeoPrism to take the images it has classified and automatically link them to
address and location data. Finally,
Geosemble will apply its Online Data Extraction agent technology to find and
apply information from publicly available “deep web” sources.
As a result, GeoPrism will automatically deliver a complete package of
images, their locations, and information linked to those images from the web.
Using these techniques, GeoPrism becomes a kind of “eye”
that can be adjusted to see various objects of interest and then can generate
highly refined, extremely targeted marketing leads for a range of users,
including manufacturers and service providers seeking to both maintain existing
residential property and associated equipment and boundaries, as well as those
looking to fill open spaces with various structures and outdoor products.
About
Geosemble Technologies:
Geosemble Technologies Inc., provides technology
for automatically and accurately fusing a range of data sources to create
decision-ready knowledge. Data
fusion sources include maps, satellite imagery, aerial imagery, road network
data, and online text and imagery. Based
in
For
further information please contact:
