欢迎来到天天文库
浏览记录
ID:31982474
大小:4.81 MB
页数:81页
时间:2019-01-30
《基于小波分形的图像处理算法.研究与应用》由会员上传分享,免费在线阅读,更多相关内容在教育资源-天天文库。
1、武汉理工大学硕士学位论文theoriesofsecondgenerationwaveletanddual-treecomplexwaveletarediscuessed,wehavealsoproposedtwonovalde-noisingmethodbasedonscal-adaptivesecondwaveletanddual.treecomplexwavelet。Finallythetestingshavedprovedtheeffectivenessofthoestwonovalmethod,andtheexperi
2、mentsshowthatthosetwonovalproposedalgorithmisefficientinimagede-noising,∞theybotllhavegoodapplicationvalueinimagede-noising。Theapplicationoffractalinimageedgedetectionofwaveletisintroducedindetail,studyingontheselectionofthebestedgewaveletbasisbasedonfraetal。Thetrad
3、itionalmethodsofedgedetectionarebasedonone-orderderivative’Smaximum,ortwo·orderderivative’Szero-crossing,andmeyareverysensitivetonoise。Inthispaper,wechoosethebestedgewaveletbasisbyusingofthefractalconstantofdifferentimages。ItWasprovedthatthischoosingedgewaveletmetho
4、disquiteeffective,theeffectofedgedetectionwouldkeepedgepositionsverywell。Aadaptivethresholdedgedetectionalgorithmofwaveletisputforwardinthispaper,andthetestingshavedprovedtheeffectivenessofthismethod,theedgeeffectisbetterthantraditionalmethodofwaveletbasedonmodulusm
5、axilnum,theexperimentalsoprovedthatthemethodproposedhaveverygoodapplicationvalueinimageedgedetection。Theapplicationofwaveletinimagefusionisstudied。Differentwaveletbasisalgorithminimagefusionandtheinfluenceoffusioneffectwhenusingdifferentdecomposinglevelsareintroduce
6、dindetail。Inthispaper,wepresentaschemeofmulti-resolutionfortwodifferentimagesusingwavelettransform。Thetwodifferentimagesweredecomposedtosomesub-bandimagesindifferentffequence,weCallprocessthedifferentcofficiencesindifferentfrequeneeandcouldgetnewwaveletcofficiences,
7、Thenthewaveletcofficiencesmatrixwastransformedinversely,Finallywecouldgetanfusedimage。Atlast,theexperimentverifiedtheeffectivenessofthisschemeandthefusioneffectwasbetter,italsoshowthatthismethodproposedhaveverygoodapplicationvalueinimagefusion·KeyWords:waveletbasis,
8、fractal,imagedenoising,edgedetection,imagefusionIV独创性声明本人声明,所呈交的论文是我个人在导师指导下进行的研究工作及取得的研究成果.尽我所知,除了文中特别加以标注和致谢的地方外,论文中不包含其他人已经发表或撰写过的研究成果,
此文档下载收益归作者所有