基于高分辨距離像的雷達目標識別技術研究.pdf

1、聲明本學位論文是我在導師的指導下取得的研究成果，盡我所知，在本學位論文中，除了加以標注和致謝的部分外，不包含其他人已經(jīng)發(fā)表或公布過的研究成果，也不包含我為獲得任何教育機構的學位或學歷而使用過的材料。與我一同工作的同事對本學位論文做出的貢獻均己在論文中作了明確的說明。研究生簽名：么泣沙婢與月弘日學位論文使用授權聲明南京理工大學有權保存本學位論文的電子和紙質文檔，可以借閱或上網(wǎng)公布本學位論文的部分或全部內容，可以向有關部門或機構送交并授權其

2、保存、借閱或上網(wǎng)公布本學位論文的部分或全部內容。對于保密論文，按保密的有關規(guī)定和程序處理。研究生簽名：么凄獨s年乃月加日AbstractModenlmtomlatlonwarpromotesthedevelopmentofradarautomaticta瑪etrecognition(M㈣，whichisbenefitfromthebetterresearchconditionsandmebetterengineefingrealizat

3、ioncreatmgbytherapiddevelopmentofmodemradarHighresolutionrangeprofile(HRI沖)hasgainedmuchattention，becauseitprovidesthedetails臼mctureinformationoftlletargetontherangedlmension，andhastheadvantageofeasyacquisitionandhandlin

4、gGrevsystemtleoryasanewwaytoanalyzeandprocesssystemswithincompleteinfo瑚ation、hasextendednsappliCationtATRrecentlyThethesisbasesonRATRapplyinggreyrelational觚a歸s，8ndtheemphasesofitareplacedontheresearchesofRATRusingHRRPbas

5、edonangleincidenceandmorphologyincidenceThemaincontentsaredescribedasfollowsFirstt0eliminatetheaspectsensitivityofHRRPRATRbasedonse舯entingangularsector8adaptively1SdiscussedThemethodmakesuseofthematchingdegreetodividethe

6、boundaryofdata也usswitchesthenon—linearclassificationproblemtoalinearone，andscane眥s廿engthdistfibufingprofiletOrepresenttheHRRPofan2LI炒larsectorItcandetermmetheangularscope，reducethenumberoftemplates，andimprovetherecogniti

7、onperformanceNextmbasedonGreySystemisintroducedTheprocessappliesgreyrelationalana上yslsonthebaslsofadaptiveangularsectorssegmentation，especiallydiscussingthean91erelatlonaldegree，animprovedgreyrelationaldegreeThethesispro

8、posesmethodofapp】vingtheanglerelmionaldegreetoRATRbasedonrealHRRPandcomplexHRRPUnlikerealH啪weshouldextractdifferentialphaseinformationofcomplexHRRPinordertotal【etheinitialphasesensitivityawayF1nally，thethesisproposesRATR

9、basedonmorphologytO眥erreducethe鋤刪ofcalclllatlonMonteCarlosimulationprovesthatthemethodcancompressthesystemstorage，ajsoenhancetherealtimeperformanceandtheantinoisepropertyaswell，mostofallitcanimprovetherecognitionperforma