基于水平集的擬合能量分割技術(shù)研究.pdf

1、華中科技大學(xué)碩士學(xué)位論文基于水平集的擬合能量分割技術(shù)研究姓名：于榮國申請學(xué)位級別：碩士專業(yè)：計算數(shù)學(xué)指導(dǎo)教師：路志宏20100528華華中科技大學(xué)碩士學(xué)位論文文IIAbstractWiththedevelopmentofdigitaltechnologymedicalimagingtechnologyhasbecomeoneofthemostimptantmeansinclinicalmedical.Indertoeffectively

2、makeuseofmedicalimageinfmationmedicalimagesegmentationtechnologyhasbeengraduallyenteredintothefieldofmedicine.Asanimptantpartofmedicalimageprocessingmedicalimagesegmentationhasacertainpracticalvalueinrespecttoseparatinge

3、xtractingthetumtargetstissues.Recentlypeoplehasproposedthoussofsegmentationmethodswithspecifictheetics.SinceLevelsetmethodwasproposedasanmathematicaltheyfeffectivelysolvingtheevolutionaryquestionsofcurvessurfaceithadgrad

4、uallybecomingaresearchfocusinthefieldofimagesegmentationisextensivelyappliedinotherprocessesofImageProcessing.ThethesisstudiestheimagesegmentationtechnologybasedonLevelsetrealizesthesegmentationalgithmsofgraycolimagesbas

5、edontwomodelssolvestheproblemsthatcan’tsegmentimageswithinconsistentintensity.Themaincontentofthisthesisisasfollows:FirstlyintroducesthemathematicalprinciplesofLevelsetmethodanalyzesonthemeritsdemeritsofthetraditionalmet

6、hoddiscussestheproblemofreinitializelevelsetfunction.SecondlydiscussestheMumfdShahmodelitssimplifiedmodel.ThroughstudyingdeeplyChanVesemodelbasedonLevelsetIrealizethesegmentationofgreyimagesbasedonthemodelpresentsthecoli

7、magessegmentationalgithm.InaccdancewiththeexperimentalresultsIconcludethemeritsofthealgithmanalysesthereasonsthatthismodelcan’tsegmentimageswithinconsistentintensityproposesthesolutions.ThirdlyintroducestheLocalexpedfitt

8、ingenergymodelbasedonLevelsetpresentsthesegmentationalgithmofgraycolimagesbasedonthismodelsolvesthedrawbackofChanVesemodel.Theexperimentalresultsshowsthisalgithmcouldsegmentpreciselyimageswithinconsistentintensitysegment