Multi-Level Sensor Fusion Algorithm Approach for BMD Interceptor Applications

Abstract

The exoatmospheric intercept regime poses a difficult technical challenge for on board interceptor discrimination, especially against mature threats that stress the seeker discrimination timeline and ability to distinguish differences between RV and decoy target characteristics. The performance may be further degraded when the observed threat characteristics are different from a priori expectations. The Discriminating Interceptor Technology Program (DITP) is being pursued by BMD0 to increase the discrimination performance robustness of NMD and TMD interceptors through fabrication and testing of advanced sensor hardware concepts and advanced sensor fusion algorithms. Advanced sensor concepts include onboard LADAR in conjunction with a multi-color passive IR sensor. Intelligent processing which can effectively combine sensor data from disparate sensors by selecting and using only the most beneficial sensor phenomenology data is a critical element of future exoatmospheric interceptor systems. The major goal of these processing algorithms is: to make optimal use of the multi-sensor data in good a priori conditions and to also provide robust discrimination when confronted with off-nominal or non a priori conditions. This paper summarizes the intelligent processing algorithms being developed, implemented and tested to intelligently fuse data from LADAR and passive infrared sensors at both the feature and decision levels. These intelligent algorithms employ dynamic selection of feature sets and the weighting of multiple classifier decisions for performance optimization, while minimizing on board processor memory and throughput requirements. Feature sets can be dynamically selected based on an estimate of the individual feature confidence. Target designation decisions can be improved by fusing weighted individual classifier decisions, the Outputs of which contain an estimate of the confidence of the data and decisions.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1998
Accession Number
ADA357702

Entities

People

  • Bart Smith
  • Charles Bjork
  • Doug Allen
  • John Rushing
  • Norman Morris

Tags

Communities of Interest

  • Sensors

DTIC Thesaurus Topics

  • Acquisition
  • Classification
  • Data Acquisition
  • Defense Systems
  • Detectors
  • Feature Extraction
  • Identification
  • Infrared Detectors
  • Laser Radar
  • Laser Sensors
  • Lethality
  • Machine Learning
  • Observation
  • Sensor Fusion
  • Target Designators
  • Test And Evaluation
  • United States

Readers

  • Computer Vision.
  • Distributed Systems and Data Platform Development
  • Missile Defense Systems.