Env aware classification in SAS for MCM

Abstract

The goal of this effort is to investigate and develop an environmentally aware and adaptive automated target recognition (ATR) system that leverages multi-frequency and multi-aspect information to autonomously characterize the seabed and appropriately tune the features, classifiers and fusion methods used within the ATR system. Algorithms to perform automated detection of underwater mine objects need to be adaptive to the environment and account for the unique characteristics of SAS imagery. This work will pursue a novel approach to the MCM problem for SAS that will investigate several techniques to characterize seabed structure and tailor detection algorithms to this environment. We will investigate the strength of standard and novel features for the characterization of the seabed and for discrimination of mine objects, utilize these features in a variety of classifiers and clustering algorithms to produce soft segmentations of the seabed, track performance of the mine detection with respect to seabed characterization, investigate environmentally aware classification, and finally to develop an operational testbed of the most promising results of MCM for use by the Navy.

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 08, 2016
Source ID
N000141612323

Entities

People

  • James F. Keller

Organizations

  • Office of Naval Research
  • United States Navy
  • University of Missouri System

Tags

Fields of Study

  • Computer science

Readers

  • Computer Vision.
  • Maritime and Naval Warfare Studies
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.