圖像邊緣檢測(cè)算法

1、青島大學(xué)畢業(yè)設(shè)計(jì)英文資料翻譯第1頁(yè)imageedgeexaminationalgithmAbstractDigitalimageprocessingtookarelativequiteyoungdisciplineisfollowingthecomputertechnologyrapiddevelopmentdaybydayobtainsthewidespreadapplication.Theedgetooktheimageonekin

2、dofbasicacteristicinthepatternrecognitiontheimagedivisiontheimageintensificationaswellastheimagecompressionsooninthedomainhasamewidespreadapplication.Imageedgedetectionmethodmanyvariedinwhichbasedonbrightnessalgithmisstu

3、diesthetimetobemostlongthetheydevelopsthematurestmethoditmainlyisthroughsomedifferenceoperatcalculatesitsgradientbasedonimagebrightnessthechangethusexaminestheedgemainlyhasRobertLaplacianSobelCannyoperatssoonLOG。Firstasa

4、wholeintroduceddigitalimageprocessingtheedgedetectionsurveyhasenumeratedseveralkindofatpresentcommonlyusededgedetectiontechnologythealgithmstwokindstouseVisualtheClanguageprogrammingrealizationthroughwithdrawstheimageres

5、ulttotwoalgithmsthecomparisontheresearchdiscussestheirgoodbadpoints.FewdInimageprocessingasabasicacteristictheedgeoftheimagewhichiswidelyusedintherecognitionsegmentationintensificationcompressoftheimageisoftenappliedtohi

6、ghleveldomain.Therearemanykindsofwaystodetecttheedge.Anywaytherearetwomaintechniques:oneisclassicmethodbasedonthegraygradeofeverypixeltheotheroneisbasedonwaveletitsmultiscaleacteristic.Thefirstmethodwhichisgotthelongestr

7、esearchgettheedgeaccdingtothevarietyofthepixelgray.ThemaintechniquesareRobertLaplaceSobelCannyLOGalgithm.ThesecondmethodwhichisbasedonwavelettransfmutilizestheLipschitzexponentacterizationofthenoisesingularsignalthenachi

8、evethegoalofremovingnoisedistillingtherealedgelines.Inrecentyearsanewkindofdetectionmethodwhichbasedonthephaseinfmationofthepixelisdeveloped.Weneedhypothesizenothingaboutimagesinadvance.Theedgeiseasytofindinfrequencydoma

9、in.It’sareliablemethod.Inchapteronewegiveanoverviewoftheimageedge.inchaptertwosomeclassicdetectionalgithmsareintroduced.Thecauseofpositionalerrisanalyzedthendiscussedameprecisionmethodinedgeientation.Inchapterthreewavele

10、ttheyisintroduced.Thedetectionmethodsbasedonsamplingwavelettransfmwhichcanextractmaimedgeoftheimageeffectivelynonsamplingwavelettransfmwhichcanremaintheoptimumspatialinfmationarerecommendedrespectively.Inthelastchapterof

11、thisthesisthealgithmbasedonphaseinfmationisintroduced.UsingthelogGabwavelettwodimensionfilterisconstructedmanykindsofedgesaredetectedincludingMachBwhichindicatesitisaoutstingbiosimulationmethod。Mayallthewkinthispaperisof

12、somevaluetoresearchapplicationsofimageedgedetection.FirstFirstchapterchapterintroductionintroduction1.11.1imageimageedgeedgeexaminationexaminationintroductionintroduction青島大學(xué)畢業(yè)設(shè)計(jì)英文資料翻譯第3頁(yè)thistothegoaldivisiontherecogniti

13、onsoonfollowingprocessingtobringtheenmousconvenience.Generallyspeakingtheabovemethodallisthewkwhichdoesbasedontheimageluminanceinfmation。Inthemultitudinousscientificresearchwkerunderhasobtainedtheverygoodeffectdiligently

14、.Butbecausetheimageedgereceivesphysicalconditionsoontheilluminationinfluencesquitetobebigaboveoftenenablesmanytohaveacommonshtcomingbasedonbrightnessedgedetectionmethodthatistheedgeisnotcontinualdoesnotsealup.Consideredt

15、hephaseinfmationintheimageimptanceaswellasitsstableacteristiccausesusingthephaseinfmationtocarryontheimageryprocessingintonewresearchtopic。Inthispapersoonintroducesonekindbasedonthephaseimageacteristicexaminationmethodph

16、aseunifmmethod.ItisnotusestheimagetheluminanceinfmationbutisitsphaseacteristicnamelysuppositionimageFouriercomponentphasemostconsistentspotachievementacteristicpoint.Notonlyitcanexaminebrightnessacteristicssoonstepacteri

17、sticlineacteristicmeovercanexamineMachbeltphenomenonwhichproducesasaresultofthehumanvisionsensationacteristic.Becausethephaseunifmitydoesnotneedtocarryonanysuppositiontotheimageacteristictypetherefeithastheverystrongvers

18、atility。1.21.2imageimageedgeedgedefinitiondefinitionTheimagemajitymaininfmationallexistsintheimageedgethemainperfmanceftheimagepartialacteristicdiscontinuityisintheimagethegradationchangequitefierceplacealsoisthesignalwh

19、ichweusuallysaidhasthestrangechangeplace。Thestrangesignalthegradationchangewhichmovestowardsalongtheedgeisfierceusuallywedividetheedgefthestepshapetheroofshapetwokindoftypes(asshowninFigure11).Inthestepedgetwosidegreylev

20、elshavetheobviouschangeButtheroofshapeedgeislocatedthegradationincreasethereducedintersectionpoint.Mayptraytheperipheralpointinmathematicsusingthegradationderivativethechangetothestepedgetheroofshapeedgeasksitsstepthesec

21、ondtimederivativeseparately。Toanedgehasthepossibilitysimultaneouslytohavethestepthelineedgeacteristic.FexampleonasurfacechangesfromaplanetothenmaldirectiondifferentanotherplanecanproducethestepedgeIfthissurfacehastheedge

22、scnerswhichtheregularreflectionacteristicalsotwoplanesfmquitetobesmooththenwksaswhenedgescnerssmoothsurfacenmalaftermirrsurfacereflectionangleasaresultoftheregularreflectioncomponentcanproducethebrightlightstripontheedge