PROJECT TITLE: Distributed Tools for Mining Remotely Sensed Data
TECHNICAL ABSTRACT (LIMIT 200 WORDS)
There is rapid growth in the amount of remotely sensed (RS) image data. All future NASA and commercial remote sensing projects will have to deal with databases of unprecedent size and complexity. Providing an effective information link between these vast data repositories and the end user is the main obstacle to building a mass market. We propose to develop GeoBrowse, an integrated prototype for interpretation, management and mining of RS data. Key innovations are: 1) support for highly distributed data visualization, analysis, archive and access functions across platforms and the Internet; 2) linking of images, and time series of images, from a variety of RS sensors with intuitive analysis, statistical modeling and prediction tools; 3) support for content-based queries, based on one or more physically relevant database attributes; 4) statistical database management functions to address database scalability, organize or cluster database attributes, visualize hidden trends and associations among attributes, and enable intelligent data mining on a regional scale. Integration of these functions over a distributed network will benefit both end user and provider. The former will access a sophisticated analysis package without purchasing it. The latter will provide a highly competitive and diversified service with superior data mining tools.
POTENTIAL COMMERCIAL APPLICATIONS
We propose the creation of an image database with linked analysis and visualization capabilities whose functionality is accessible from the world wide web. This allows for the utility of remotely sensed data to be extended into communities that otherwise would not have the infrastructure and knowledge to benefit from NASA"s MTPE vast data resources. The commercial opportunities potentially lead to a broad range of users from application areas such as precision crop monitoring, forestry, transportation, land development and real estate, mineral exploration, and the environment. The underlying technology is readily usable in other application areas including medical imaging, commercial data mining and multi-media applications. The research will also make substantial contributions to MathSoft"s core products on data mining.
NAME AND ADDRESS OF PRINCIPAL INVESTIGATOR
Giovanni B. Marchisio
MathSoft, Inc. Data Analysis Product Div
1700 Westlake Ave. N. Suite 500
Seattle , wa 98109
NAME AND ADDRESS OF OFFEROR
MathSoft, Inc. Data Analysis Product Div
1700 Westlake Ave. N. Suite 500
Seattle , WA 98109