改进的mumford-shah模型与其在医学图像处理中的应用-(5885)

改进的mumford-shah模型与其在医学图像处理中的应用-(5885)

ID:33005869

大小:1.68 MB

页数:48页

时间:2019-02-19

改进的mumford-shah模型与其在医学图像处理中的应用-(5885)_第1页
改进的mumford-shah模型与其在医学图像处理中的应用-(5885)_第2页
改进的mumford-shah模型与其在医学图像处理中的应用-(5885)_第3页
改进的mumford-shah模型与其在医学图像处理中的应用-(5885)_第4页
改进的mumford-shah模型与其在医学图像处理中的应用-(5885)_第5页
资源描述:

《改进的mumford-shah模型与其在医学图像处理中的应用-(5885)》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库

1、ABSTRACTedgesoftheimagethathaveblurrededgesaccurately.Inordertoovercometheaboveshortcomings,weproposeafastimprovedhierarchicalmultiphaselevel—setsegmentationmodelwhichbasedonthehierarchicalmultiphasesegmentationmodel,fromtheexpectsofthemeancurvature,level-setf

2、unction①andDiracfunction万(①)jntheCHIVeevolutionequationsandallintroductionofananisotropicdiffusionequation,wemodifytheC—Vmodelandgetourimprovedhierarchicalmultiphaseimagesegmentationmodel.Experimentalresultssuggestthatourmodelismoreefficientandfasterinsegmenta

3、tionofmulti-objectimageandobjectswithweakboundaries.TheC-Vmodelalsohasmanydisadvantages:1.Lackofpriorknowledgeandstructureinformationofthedatamodel;2.Thesegmentatedresultsdependonthechoiceofinitialconditions;3.Unreasonablechoiceofweightfactorswillleadtobadsegm

4、entationresults,ifweonlyuseexpedencetodecidethevaluesoftheweightfactors,itwilllowertheuniversalityandautoprocessingabilityofthealgorithm.TheGaussianmixturemodelisaprobabilitymodelwhichapproachstheimagehistogram,also,itisaidealmodelwhichcandescribetheslowvaryin

5、gofthegraylevelintheregionsandthecharacterofthewholeimage,SOweintroducethismodeltodescribetheimage,andusetheposterprobabilitytomodifytheC-Vmodel.Forwedothis,wehavethreeobjectives:(1)noneedtosettingdifferentparameterfortheevolutionequationfordifferentimages.(2)

6、thesinglecurvesettingatanypositionscansegmenttheobjectsweareinterestedin.(3)wecansegmenttheimagesaccuratelyattheedge.C—Vmodelalsotrapsintolocalsolutionwhenitsegmenttheimagesatthesinglescale.Multi-resolutionmethodsobtainaglobalviewofanimagebyexaminingitatvariou

7、sresolutionlevels,andperfectlyunifythecontradictionbetweentheaccuracyinthehigherresolutionandtheeasinessforsegmentationinreducedresolution,thenitavoidsthelocalminimum.C-Vmodelalsolackspriorknowledgeandstructuralinformationofthedatamodel,butMarkovrandomfield(MR

8、F)modelsprovideapowerfulandformalwaytoaccountforspatialdependenciesbetweenimagepixels.So,weconnectthemultiresolution,MRF,andC-Vmodeltogether,andproposeanewhybridmodel,andthisnewmod

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。