---------- NIMA ----------

2 Phase I Selections from the 01.1 Solicitation

(In Topic Number Order)
352-C Christopher Avenue
Gaithersburg, MD 20879
(301) 990-1100
Dr. Thomas Tsao
NIMA 01-001      Selected for Award
Title:Attention Guided Pattern and Geospatial Analysis System
Abstract:The goal of this project is to develop an attention-guided pattern and geospatial analysis system for imagery exploitation. This new technology will be based on the dynamic receptive field model of visual attention that has emerged from neuroscience research. The attention system holds the key to the success of the vision systems of many higher form animals struggling to survive: They see what they attend to. What is the attention system and why is it so powerful? We propose to develop a computational theory of attention and apply this theory to build an image exploitation system that can perform feature extraction, object recognition, feature registration, change detection, and related image geospatial analysis tasks.The development of this attention-guided pattern and geospatial analysis system will provide solutions to many urgent technical problems in imagery intelligence and geospatial analysis. In particular, the capability to rapidly match an invariant pattern to sensory input with myriad possible variations will have a revolutionary impact on imaging sensor technology and autonomous system design.

4676 Admiralty Way, 10th floor
Marina del Rey, CA 90292
(310) 448-8275
Dr. Steven Minton
NIMA 01-002      Selected for Award
Title:Constraint-based User Interface for Integrating Geospatial Data Sources
Abstract:With the large amount of geospatially-referenced material available, an important problem for NIMA is how to organize and present that information to their customers. We propose to address this problem by building a geospatial catalog system that will not just tell you which products are available, but also allow the user to see the actual products and the relationship to other products. This will allow the NIMA customers to determine exactly which products contain the information they need and then order those products from NIMA. One of the key pieces of technology that we previously developed is a constraint-based reasoning system that ties the various types of information together. The key advance provided by this tool is the ability to provide a coordinated view of a wide variety of heterogeneous geospatial data. We propose to take this initial system and apply it directly to the problem of providing a universal graphic geospatial interface for all of NIMA's data sources.The technology described in this proposal will form the basis of a new product for integrating geo-spatial information. This product will enable geospatial information to be flexibly integrated for application-domains (like those described this proposal) where multiple sources of information must be integrated with maps and/or satellite images. Potential customers, in addition to U.S. military and intelligence services, include insurance companies, environmental organizations, emergency response/aid organizations, news organizations, surveyors, mining and oil companies, timber harvesters, and even film companies (for location management). And, in addition to the commercial benefits from this product, the research will push forward the fundamental technology we are developing for flexibly integrating heterogeneous geospatial information from multiple sources.