Interpretation and Fusion of Sonar Images Using Evidential Reasoning

Abstract

In mine warfare classification of bottom objects is essential. It is also one of the major problems. Especially when there is considerable sea-bottom clutter as in many MCM situations. This report presents a novel approach to automatic interpretation of sonar images of the sea-bottom, which allows combination of information from multiple images of the same site.The method provides increased reliability in the interpretation compared to existing methods, as information from several images is utilized. Additionally, a distinction is made between uncertainty and ignorance. This distinction is relevant as imprecision in an interpretation of a sonar image can be due to both ambiguity and to lack of knowledge. The viability of the technique is demonstrated by combining sonar images acquired with an experimental sonar with two simultaneous views.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1995
Accession Number
AD1114016

Entities

People

  • B. Stage

Organizations

  • SACLANT ASW Research Centre

Tags

Communities of Interest

  • Autonomy
  • Sensors

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Artificial Intelligence Computing
  • Artificial Intelligence Software
  • Bayesian Networks
  • Computer Vision
  • Databases
  • Detection
  • Expert Systems
  • Image Processing
  • Images
  • Mathematical Models
  • Models
  • Nato
  • Neural Networks
  • Object Recognition
  • Probabilistic Models
  • Probability
  • Recognition
  • Reconnaissance
  • Sonar Images
  • Surveillance
  • Underwater Vehicles

Readers

  • Acoustical Oceanography.
  • Artificial Intelligence
  • Computer Vision.