NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Back to Results
Using AGNESS (A Generalized Network-based Expert System Shell) for matching imagesThe image correspondence problem is considered the most difficult step in both stereo and motion analysis. Stereo vision is useful in determining the 3-D positions of points on visible surface in a scene. Motion analysis is useful in determining the spatial and temporal relationships of objects in an environment. Besides stereo and motion analysis, there is the image correspondence problem. Most of this work is based on point or local area properties of the observed gray level values in 2-D images. A global and general approach to this problem is described by using a knowledge-based system. The knowledge it uses consists of both physical properties and spatial relationships of the edges and regions extracted from the given images. The physical component depends on features of the edge or region) in isolation. The spatial component involves the set of edges and regions adjacent to a given edge (or region) of the first image and the set of edges and regions adjacent to each potentially matching edge (or region) of the second image; thus the spatial context of each edge or region is considered. A computational network is used to represent this knowledge, it allows the computation of the likelihood of matching two edges or regions with logical and heuristic operators. An expert system shell called AGNESS (A Generalized Network-based Expert System Shell) is used to build a prototype system.
Document ID
19890006210
Acquisition Source
Legacy CDMS
Document Type
Conference Paper
Authors
Pong, Ting-Chuen
(Minnesota Univ. Minneapolis, MN, United States)
Lee, Chung-Mong
(Minnesota Univ. Minneapolis, MN, United States)
Slagle, James
(Minnesota Univ. Minneapolis, MN, United States)
Date Acquired
September 5, 2013
Publication Date
October 1, 1988
Publication Information
Publication: NASA, Marshall Space Flight Center, Fourth Conference on Artificial Intelligence for Space Applications
Subject Category
Computer Programming And Software
Accession Number
89N15581
Distribution Limits
Public
Copyright
Work of the US Gov. Public Use Permitted.

Available Downloads

There are no available downloads for this record.
No Preview Available