Application of Signal Analysis and Pattern Recognition Procedures to Study Blast Induced Ground Motion.

Abstract

The second-year results of an effort to introduce computer automated signal analysis and pattern recognition procedures to the study of blast-induced ground motion is presented. Intermediate results involving the characterization of ground motion acceleration records to identify ground spall versus no spall regions is reported. A computerized procedure extracted information from the time, frequency, and cepstrum description of the signal. These signal features were then incorporated into a Fisher's Linear Discriminant (FLD) pattern recognition procedure. Procedures to identify good and bad features are reviewed. This identification seems to be required for the efficient and effective application of the FLD. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1984
Accession Number
ADA143744

Entities

People

  • J. M. Carson

Tags

Communities of Interest

  • Counter IED
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Accelerometers
  • Air Force
  • Air Force Facilities
  • Air Pressure
  • Computer Programs
  • Computers
  • Engineering
  • Feature Extraction
  • Frequency
  • Gray Zone
  • Identification
  • New Mexico
  • Pattern Recognition
  • Recognition
  • Scientific Research
  • Signal Processing
  • Test Sets

Readers

  • Regression Analysis.
  • Seismology
  • Speech Processing/Speech Recognition.

Technology Areas

  • AI & ML