Accession Number : AD0759554

Title :   Mathematical Research in Target Detection and Identification,

Corporate Author : MCDONNELL DOUGLAS ASTRONAUTICS CO-WEST HUNTINGTON BEACH CALIF

Personal Author(s) : Parkyn,W. A. ; Webber,W. F. ; Webb,S. R.

Report Date : APR 1973

Pagination or Media Count : 35

Abstract : The report describes research results on mathematical techniques useful for processing signal and image data for the purpose of target detection and identification. Faster algorithms have been developed for calculating the correlation function for two-dimensional data. Unlike the standard approach which use the cyclic Fourier transform, this approach is based on Walsh functions and is available in both cyclic and non-cyclic versions. The potential for using Walsh transforms for deconvolution is discussed. Finite impulse response filters are useful for estimating a two-dimensional image in the presence of noise. An improved version of such filters is presented. Techniques for calculating the discrete Fourier transform of two-dimensional real data are discussed, which offer computational advantages over the standard method for complex data. Finally, the results of an eigenvector expansion technique for analyzing nonstationary signals is presented with applications to acoustic data. (Author)

Descriptors :   (*INFORMATION THEORY, TRANSFORMATIONS(MATHEMATICS)), (*TARGET RECOGNITION, ACOUSTIC DETECTORS), TARGET DISCRIMINATION, IDENTIFICATION SYSTEMS, CORRELATION TECHNIQUES, SAMPLING, MATRICES(MATHEMATICS), PATTERN RECOGNITION, TIME SERIES ANALYSIS

Subject Categories : Statistics and Probability
      Cybernetics
      Acoustic Detection and Detectors

Distribution Statement : APPROVED FOR PUBLIC RELEASE