Project Title:

Symbolic Imagery Management System

07.08-8430
Symbolic Imagery Management System 
Consultant's Choice, Inc.
8800 Roswell Road #130
Atlanta, GA 30350
Paul D. Lampru, Jr. (404-992-8430)
GSFC -- NAS5-30271

Abstract:

	 The innovation pursued in this project is a symbolic, imagery-management system
(SIMS) for use initially by NASA scientists and, subsequently, by remote-sensor scientists
external to NASA. Integrating image data with graphic data structures and feature
information is very difficult using traditional digital representations. Separate
image and feature information data bases are commonly used, but the analyst is required
to maintain two or more databases and to perform mentally the integration of the
information. This project applies an innovative digital-to-symbolic transformation
algorithm (DSTA) to NASA-supplied imagery. This innovation employs a technique that
represents gray-scale images symbolically in Lisp to develop software that allows
smooth integration of image data, graphic data structures, and feature information.
The user will be able to create and maintain such an integrated database, to embed
feature information directly into images, to integrate graphic data structures with
images, and to query the image database based upon information content.
	Phase I successfully demonstrated the feasibility and potential of both DSTA and
symbolic image-analysis and, in addition, found that the use of artificial neural
systems provides a substantive increase in the efficiency and accuracy of image segmentation
and classification .

Potential Commercial Application:

Potential Commercial Application: Commercial applications may be for satellite imagery
and imagery dealing with manufacturing productivity and robotics.