AN APPLICATION OF PATTERN RECOGNITION TO RADIOMETRIC TARGET DETECTION.

Abstract

Classical detection theory is used to provide a framework for the study of the potential of passive detection of metallic targets by millimeter wave radiometry. The target is assumed to be embedded in a foliage environment. The problem is characterized as a two-class detection problem. Class C sub 1 denotes the class of measurements obtained when the field of view V of the radiometer contains some target elements, and C sub 2 represents the class of measurements obtained when V contains no target elements. Each of the measurement sets is characterized by probability density functions. These functions are used to obtain operating characteristic curves relating alpha and beta errors and to determine discriminant functions for the detection problem. The alpha error is the probability of assigning an observation to class C sub 1 when it belongs to C sub 2, and the beta error is the probability of assigning an observation to class C sub 2 when it belongs to C sub 1. Operating characteristic curves are useful in determining the amount of target obscuration for various alpha and beta errors. (Author)

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1969
Accession Number
AD0684904

Entities

People

  • Richard A. Mcgee
  • William G. Lese Jr.
  • William J. Sacco

Organizations

  • Ballistic Research Laboratory

Tags

DTIC Thesaurus Topics

  • Detection
  • Measurement
  • Millimeter Waves
  • Observation
  • Pattern Recognition
  • Probability
  • Probability Density Functions
  • Radiometers
  • Radiometry
  • Recognition
  • Target Detection

Readers

  • Computer Vision.
  • Radar Systems Engineering.
  • Statistical inference.

Technology Areas

  • 5G
  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms