Exam (elaborations) Multichannel 2-D

SlTv CLASSifiCATlON O* Tmi? PaGE REPORT DOCUMENTATION PAGE i REPORT SECURITY Classification UNCLASSIFIED lb RESTRICTIVE MARKINGS i SECURITY Classification authority 0EClASSi*iCATiON /DOWNGRADING SChEOULE ) DISTRIBUTION/ AVAILA8H1TY Of REPORT Approved for public release; distribution is unlimited PERFORMING ORGANISATION REPORT NUM8ER(S) S MONITORING ORGANISATION REPORT NUM8ERIS) NAME OF PERFORMING ORGANIZATION Naval Postgraduate School 60 OFMCE SYMBOL (It JOOiXtO'*) 62 7* NAME OF MONITORING ORGANISATION Naval Postgraduate School AOORESS ;C/y. Stitt. *nd HP Cod*) Monterey, California 7b AOORESS (Cry. Suf*. »ndZiPCod*) Monterey, California NAME OF FuNOlNG / SPONSORING ORGANISATION 8b OFFICE SYMBOL (it *ppti(*bi*) 9 PROCUREMENT INSTRUMENT lOEN T.fiCATlON NUMBER AOORESS (Cry. Sfjf* *nd Zip Cod*) 10 SOURCE OF FUNDING NUMBERS PROGRAM ELEMENT NO PROJECT NO TAS NO WORK jNlT ACCESSION NO t.TlE (include S*cu"ry CUmtiation) MULTICHANNEL 2-D POWER SPECTRAL ESTIMATION AND APPLICATIONS P£3SONA k auThOR(S) El-Shaer. Hamdv T.M. TySi OF REPORT Ph.D. Dissertation i 3b T'ME COvtREO FROM TO 14 DATE OF REPORT (Yt*r Month Q*y) 1987 November IS PAGE COuNT 304 Supplementary notation COSATi COOES ' ElD GROUP Subgroup '8 SUBJECT TERMS (Continue on revert* it necemry »nd identify by O/OC* number) Signal Modeling Linear Prediction Image Coding AR Models Maximum Likelihood Method Spectral Estimation ABSTRACT (Continue on revene it netemry *nd iO*ntity by O/OK* numo*r) Spectral estimation for multiple 2-D signals by model-based methods is developed. The rocedures compute the entire spectral matrix of autospectra and cross spectra for the set of 2- ) signals. Spectral analysis by autoregressive (AR) modeling is studied extensively. Specific lifferences between AR models for this problem and those for lower dimensional problems are lighlighted. An extension of the Jackson-Chien method for combining estimates with single luadrant support is proposed and a method is developed for estimating the model parameters lirectly from the data (i.e. without prior estimation of a correlation matrix). A measure of the imilarity of two spectral estimates based on the statistical divergence is proposed and used to ompare various spectral estimates. A comprehensive set of experimental studies are presented 3 S"R'3uTiON I AVAlLAdlL'TY OF ABSTRACT 3^ClASSiFiEOAjNl'MIT£D O SAME AS RPT OTiC USERS 21 ABSTRACT SECURITY ClASSiFiGA HON UNCLASSIFIED NAVE OF RESPONSIBLE NOiviOUAl Prof. C.W. Therrien 220 TELEPHONE (irxlud* AretCod*) (408) 646-2032 22c OFFitfc SYMBOL 62Ti ORM 1473,84 mar 83 APR edition **y 0* ui*d jnt'l «in*utted All otntr edition* *'« obtoittt SECURITY Classification Q( t h .s pace UNCLASSIFIED UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE (Whan Dmm t-v»»t.0 showing the performance of the methods in estimating the autospectra and magnitude and phase of the cross spectra. The Maximum Likelihood Method (MLM) of spectral estimation is extended to the multichann