Radar Signature Investigation. Volume II. Target Classification.

Abstract

The results of a comprehensive investigation of the use of short-pulse (sub-nanosecond) radar signature characteristics for classifying radar targets are presented. Properties of an Equivalence Class Technique (ECT) are discussed relative to the preprocessing, or collapsing, of wideband signature data. Results of a discriminant investigation are presented in order to illustrate the relative merit of selected signature features for the purpose of classification. The results of classifying ten 'unknown' aerospace-type radar targets in terms of their similarity to a set of known targets are presented. The appendices of this report contain detailed discussions concerning the design, development, and evaluation of a trainable target classifier or pattern recognizer. The training procedure is explained through the use of examples. Numerous detailed examples are also provided to illustrate the use of this classifier for the purpose of classifying unknown targets. The time required for operation of the classifier is discussed and is shown to be compatible with operational real-time classification. (Author)

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 1970
Accession Number
AD0878304

Entities

People

  • Dave O. Cisco
  • George P. Coble
  • George W. Gruver

Organizations

  • General Dynamics

Tags

DTIC Thesaurus Topics

  • Classification
  • Machine Learning
  • Nanosecond Time
  • Preprocessing
  • Radar Signatures
  • Radar Targets
  • Target Classification
  • Targets
  • Test And Evaluation
  • Training

Readers

  • Business Analytics
  • Sensor Fusion and Tracking Systems.
  • Theoretical Analysis.

Technology Areas

  • Space
  • Space - Space Objects