The Application of Pattern Recognition Techniques for Tunnel Location.

Abstract

The location of buried tunnels is part of an ongoing program at DARPA with many different disciplines and techniques involved in the effort. Geo-Centers, Inc. effort to locate buried tunnels was originally funded by DARPA for a period of five months beginning in June 1976. The major thrust of that effort was the theoretical investigation of special computer assisted detection methods to locate tunnels. The basic electromagnetic signal was theoretically generated using the principles of video pulse radar. This radar technique uses a short time limited electromagnetic pulse radiated by a movable antenna platform at the earth-air interface. The major emphasis of the Geo-Centers' program was the development of a signal processing method to detect buried tunnels using video pulse radar probes. Video pulse radar is used to transmit and receive a short unidirectional electromagnetic pulse. The return pulse in turn produces a two dimensional sounding or picture. Theoretical soundings were generated based on a simple scattering model relevant to long cylindrical tunnels that are buried in granite. A fast and sensitive picture processing algorithm was evolved for the specific problem of subsurface tunnel detection.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 30, 1976
Accession Number
ADA040900

Tags

Communities of Interest

  • Air Platforms
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Detection
  • Diameters
  • Frequency
  • Geometry
  • Gray Scale
  • Materials
  • Pattern Recognition
  • Recognition
  • Scattering
  • Signal Detection
  • Signal Processing
  • Steady State
  • Targets
  • Three Dimensional
  • Time Domain
  • Transfer Functions
  • Two Dimensional

Fields of Study

  • Engineering
  • Physics

Readers

  • Electromagnetic Wave Scattering and Antenna Radiation Engineering
  • Radar Systems Engineering.
  • Technical Research and Report Writing.

Technology Areas

  • AI & ML