Accession Number : ADA319025

Title :   Neural Networks for Automatic Recognition of Military Deployment Templates.

Descriptive Note : Technical rept.,

Corporate Author : DEFENCE SCIENCE AND TECHNOLOGY ORGANIZATION CANBERRA (AUSTRALIA)

Personal Author(s) : Price, Richard ; Kerry, Katrina

PDF Url : ADA319025

Report Date : AUG 1996

Pagination or Media Count : 27

Abstract : A key function in tactical intelligence, is the ability to recognize patterns of behaviour in opposing force activities. Deployment templates are patterns that define the spatial layout opposing force units are known to employ, when carrying out a range of different operations. The early detection of these patterns, provide key indicators to intelligence analysts, for predicting future opposing force activities. In this paper, we investigate the use of neural networks to automatically detect deployment templates, when given full or partial opposing unit positional information. The spatial template recognition problem is a subset of many spatial pattern recognition problems found throughout defence, and the demonstrated success of this approach, could have important military implications beyond this particular application.

Descriptors :   *DEPLOYMENT, *NEURAL NETS, *MILITARY FORCES(FOREIGN), *PREDICTIONS, *TEMPLATES, *PATTERN RECOGNITION, *TACTICAL INTELLIGENCE, SPATIAL DISTRIBUTION, POSITION(LOCATION), DETECTION, ENEMY, AUSTRALIA, AUTOMATIC, INDICATORS, ANALYSTS.

Subject Categories : Cybernetics
      Military Intelligence
      Military Forces and Organizations

Distribution Statement : APPROVED FOR PUBLIC RELEASE