Accession Number : ADA327320

Title :   Image Understanding for Database Query.

Descriptive Note : Final rept. Sep 94-Sep 96,

Corporate Author : PAR GOVERNMENT SYSTEMS CORP NEW HARTFORD NY

Personal Author(s) : Ferrara, Charles ; Barrett, Scott

PDF Url : ADA327320

Report Date : APR 1997

Pagination or Media Count : 22

Abstract : The objective of the Image Understanding for Database Query (IU4DBQ) effort was to combine the powerful reasoning capabilities of the Loom knowledge representation system with the extensive image processing and feature extraction capabilities of KB Vision TM. Another important aspect of this effort was to apply the algorithms developed to realistic reconnaissance imagery. A major challenge was to integrate Loom and KB Vision TM within a useful context for the softcopy reconnaissance imagery available in the Rome Laboratory (RL) IE 2000 facility. The most practical use for Loom is to query the imagery server databases and determine images that have specific characteristics that are most suitable for processing. This suitability is based on the target area of interest, the presence of unconfirmed targets, and the quality of the imagery.

Descriptors :   *DATA BASES, *IMAGE PROCESSING, *INTERROGATION, ALGORITHMS, REASONING, VISION.

Subject Categories : Computer Systems
      Cybernetics

Distribution Statement : APPROVED FOR PUBLIC RELEASE