A Quaternion Framework for Color Image Smoothing and Segmentation

A Quaternion Framework for Color Image Smoothing and Segmentation

ID:39776579

大小:2.29 MB

页数:18页

时间:2019-07-11

A Quaternion Framework for Color Image Smoothing and Segmentation _第1页
A Quaternion Framework for Color Image Smoothing and Segmentation _第2页
A Quaternion Framework for Color Image Smoothing and Segmentation _第3页
A Quaternion Framework for Color Image Smoothing and Segmentation _第4页
A Quaternion Framework for Color Image Smoothing and Segmentation _第5页
资源描述:

《A Quaternion Framework for Color Image Smoothing and Segmentation 》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、IntJComputVis(2011)91:233250DOI10.1007/s11263-010-0388-9AQuaternionFrameworkforColorImageSmoothingandSegmentationÖzlemN.Subakan·BabaC.VemuriReceived:3July2009/Accepted:16September2010/Publishedonline:18November2010©SpringerScience+BusinessMedia,LLC2010AbstractInthispaper,wepresentfeature/detailpre

2、serving1IntroductionmodelsforcolorimagesmoothingandsegmentationusingtheHamiltonianquaternionframework.First,weintroduceColorconveysessentialinformationthatcanbeemployedinanovelquaternionicGaborfilter(QGF)whichcancombinemanyvisiontasksincludingbutnotlimitedtoobjectrecog-thecolorchannelsandtheorienta

3、tionsintheimageplane.nition,tracking,segmentation,registrationetc.Withsignif-Weshowthatthesefiltersareoptimallylocalizedbothintheicantadvancesincomputingpowerandmemoryoverthespatialandfrequencydomainsandprovideagoodapprox-pasttwodecades,colorimageprocessinghasattractedmuchimationtoquaternionicquadr

4、aturefilters.UsingtheQGFs,interestintherecentpast.Inthisarea,colorimagedenois-weextractthelocalorientationinformationinthecolorim-ingandsegmentationarestillelusivechallenges.Duetotheages.Second,inordertomodelthisderivedorientationin-multichannelnatureofthecolorimages,thekeyissueishowtocoupletheinfo

5、rmationcontainedinthegivencolorformation,weproposecontinuousmixturesofappropriate(e.g.red,greenandblue)channels.Consideringeachindi-exponentialbasisfunctionsandderiveanalyticexpressionsvidualchannelofacolorimageasaseparatemonochromeforthesemodels.Theseanalyticexpressionstaketheformimage,theearlyap

6、proachesofteninvolvedcomponent-wiseofspatiallyvaryingkernelswhich,whenconvolvedwithaapplicationofthetraditionalgrayleveltechniquesoneachcolorimageorthesigneddistancefunctionofanevolvingchannelseparately.However,thisapproachfailstocapturecontour(placedinthecolorimage),yieldadetailpreservingtheinher

7、entcorrelationbetweenthecomponentsandresultssmoothingandsegmentation,respectively.Severalexamplesincolorartifactsorblending.Toavoidthis,colorimageonwidelyusedimagedatabasesareshowntodepicttheper-processingshouldb

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。