畢業(yè)論文外文翻譯-數(shù)字圖像處理

1、浙江科技學(xué)院畢業(yè)設(shè)計(jì)（外文翻譯）1DigitalImageProcessing1IntroductionManyoperatshavebeenproposedfpresentingaconnectedcomponentnadigitalimagebyareducedamountofdatasimpliedshape.Ingeneralwehavetostatethatthedevelopmentchoicemodi_cationofsu

2、chalgithmsinpracticalapplicationsaredomaintaskdependentthereisnobestmethod“.Howeveritisinterestingtonotethatthereareseveralequivalencesbetweenpublishedmethodsnotionsacterizingsuchequivalencesdi_erencesshouldbeusefultocat

3、egizethebroaddiversityofpublishedmethodsfskeletonization.Discussingequivalencesisamainintentionofthisrept.1.1CategiesofMethodsOneclassofshapereductionoperatsisbasedondistancetransfms.Adistanceskeletonisasubsetofpointsofa

4、givencomponentsuchthateverypointofthissubsetrepresentsthecenterofamaximaldisc(labeledwiththeradiusofthisdisc)containedinthegivencomponent.Asanexampleinthis_rstclassofoperatsthisreptdiscussesonemethodfcalculatingadistance

5、skeletonusingthed4distancefunctionwhichisappropriatetodigitizedpictures.Asecondclassofoperatsproducesmediancenterlinesofthedigitalobjectinanoniterativeway.Nmallysuchoperatslocatecriticalpoints_rstcalculateaspeci_edpathth

6、roughtheobjectbyconnectingthesepoints.Thethirdclassofoperatsisacterizedbyiterativethinning.HisticallyListing[10]usedalreadyin1862thetermlinearskeletonftheresultofacontinuousdefmationofthefrontierofaconnectedsubsetofaEucl

7、ideanspacewithoutchangingtheconnectivityoftheiginalsetuntilonlyasetoflinespointsremains.Manyalgithmsinimageanalysisarebasedonthisgeneralconceptofthinning.Thegoalisacalculationofacteristicpropertiesofdigitalobjectswhichar

8、enotrelatedtosizequantity.Methodsshouldbeindependentfromthepositionofasetintheplanespacegridresolution(fdigitizingthisset)theshapecomplexityofthegivenset.Intheliteraturethetermthinning“isnotusedinauniqueinterpretationbes

9、idesthatitalwaysdenotesaconnectivitypreservingreductionoperationappliedtodigitalimagesinvolvingiterationsoftransfmationsofspeci_edcontour浙江科技學(xué)院畢業(yè)設(shè)計(jì)（外文翻譯）3Twopixellocationspqinthegridcellmodelarecalled0adjacenti_p6=qtheys

10、hareatleastonevertex(whichisa0cell).Notethatthisspeci_es8adjacencyin2D26adjacencyin3Difthegridpointmodelisused.Twopixellocationspqinthegridcellmodelarecalled1adjacenti_p6=qtheyshareatleastoneedge(whichisa1cell).Notethatt

11、hisspeci_es4adjacencyin2D18adjacencyin3Difthegridpointmodelisused.Finallytwo3Dpixellocationspqinthegridcellmodelarecalled2adjacenti_p6=qtheyshareatleastoneface(whichisa2cell).Notethatthisspeci_es6adjacencyifthegridpointm

12、odelisused.AnyoftheseadjacencyrelationsA__2f012461826gisirreexivesymmetriconanimagecarrierC.The_neighbhoodN_(p)ofapixellocationpincludespits_adjacentpixellocations.Codinatesof2Dgridpointsaredenotedby(ij)with1_i_n1_j_mija

13、reintegersnmarethenumbersofrowscolumnsofC.In3Dweuseintegercodinates(ijk).Basedonneighbhoodrelationswede_neconnectednessasusual:twopointspq2Care_connectedwithrespecttoM_CneighbhoodrelationN_i_thereisasequenceofpointsp=p0p

14、1p2:::pn=qsuchthatpiisan_neighbofpi?1f1_i_nallpointsonthissequenceareeitherinMallinthecomplementofM.AsubsetM_Cofanimagecarrieriscalled_connectedi_MisnotemptyallpointsinMarepairwise_connectedwithrespecttosetM.An_component

15、ofasubsetSofCisamaximal_connectedsubsetofS.Thestudyofconnectivityindigitalimageshasbeenintroducedin[15].ItfollowsthatanysethIiconsistsofanumberof_components.Incaseofthegridcellmodelacomponentistheunionofclosedsquares(2Dc

16、ase)closedcubes(3Dcase).Theboundaryofa2cellistheunionofitsfouredgestheboundaryofa3cellistheunionofitssixfaces.Fpracticalpurposesitiseasytouseneighbhoodoperations(calledlocaloperations)onadigitalimageIwhichde_neavalueatp2

17、CinthetransfmedimagebasedonpixelvaluesinIatp2CitsimmediateneighbsinN_(p).2NoniterativeAlgithmsNoniterativealgithmsdeliversubsetsofcomponentsinspeciedscanderswithouttestingconnectivitypreservationinanumberofiterations.Int