单幅图像去雨雪的算法分析

《单幅图像去雨雪的算法分析》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、ABSTRACTpartcouldbemodifiedasanon-rainornon-snowguidanceimage，whilethehighfrequencypartistreatedasaninputimageoftheguidedfilter,SOthatanon。raincomponentofthehighfrequencypartCanbeobtained．Andthenaddittothelowfrequencyparttogettherecoveredimage．Furtherwetaketheminimizationb

2、etweentheinputimageandtherecoveredimagetorestorethevalleyedges．Finallyweuseguidedfilteronceagaintogetthefinalrefinedrecoveredimage．3．AguidedL0smoothingfilterisdesignedinspiredbypreviousL0gradientminimization．BaseonthetheoryofL0gradientminimizationandexperimentalanalyzes，we

3、designaguidedLosmoothingfilter．ThedesignedfilterCansmooththeinputimageaccordingtothegradientmagnitudeofguidedimage，butnotaccordingtothestructureofguidedimagelikepreviousguidedfilter．Andfromtheexperimentalresults，thedesignedguidedL0smoothingfilterhasabetterperformancethangu

4、idedfilterinthegivencase．4．SingleimagerainandsnowremovalmethodviaguidedL0smoothingfilterisproposed．Andtheinputsofthedesignedguidedksmoothingfilterarethenon‘rainornon．snowbutblurredguidanceimageandobservedimage，whiletheoutputisacoarseno．rainorno．snowimage．Becausethelowfrequ

5、encypartlosttheinformationofvalleyedges，wetaketheminimizationbetweentheinputimageandthecoarseno．．rainorno．．snowimagetoacquirethecorrespondingfinalresult．Theproposedmethodhasbetterperceptualqualityagainsttherainandsnowremovalmethodusingmulti-guidedfilter．KeyWords：singleimag

6、erainandsnowremoval；guidedfilter；L0gradientminimization；guidedLosmoothingfilter目录第一章绪论⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯11．1课题研究背景和意义⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯．．11．2国内外的研究现状⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯。21．2．1基于视频图像的去雨雪的方法⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯．．21．2．2基于单幅图像的去雨雪的方法⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯．．31．3论文的主要研究内容

7、和创新点⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯51．4论文的结构安排⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯6第二章背景知识⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯．82．1雨雪的物理特性⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯82．1．1雨雪的外表特性⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯一82．1．2雨雪天成像的物理模型⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯．．82．2引导滤波⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯92．2．1引导滤波的定义⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯

8、⋯⋯⋯一92．2．2运用引导滤波的单幅图像去雨雪方法⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯112．3Lo梯度最小化⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯132．4本章小结⋯⋯