NUMERIC AND SYMBOLIC PROCESSING FOR SONAR INFORMATION MANAGEMENT

PDF

Authors
  1. Milios, E.
  2. Jenkin, M.
  3. Lu, F.
Corporate Authors
Defence Research Establishment Atlantic, Dartmouth NS (CAN);York Univ, North York ONT (CAN) Dept of Computer Science
Abstract
Sonar information management concerns the synthesis of high-level tactical pictures from data generated by acoustic sensors and includes the processes of data association, target motion analysis, and classification. The objective of this contract is to propose candidate numeric and symbolic processing techniques that could be applied to the problems of data integration and automation within the context of sonar information management. A study has been made of candidate techniques from the computer vision, knowledge-based signal processing and neural networks fields. Complex systems, which address tasks such as road following, temporal signal understanding, object recognition, and signal tracking, have been developed in these fields. Many of these techniques may find potential application in sonar information management. Techniques that have been identified include neural networks for signal tracking, knowledge-based signal processing architectures for introducing symbolic feedback between sonar information management processes, and the use of qualitative constraints for associating signals between sonobuoys.
Keywords
Sonar information management;Neural networks;Data association;Knowledge Based Systems
Report Number
DREA-CR-94-420 — Contractor Report
Date of publication
01 Jul 1994
Number of Pages
36
DSTKIM No
97-02198
CANDIS No
502406
Format(s):
Hardcopy;Document Image stored on Optical Disk

Permanent link

Document 1 of 1

Date modified: