基于模糊形态学的图像边缘轮廓提取分割算法.pdf

《基于模糊形态学的图像边缘轮廓提取分割算法.pdf》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、机床与液压Sep．2013HydromechatronicsEngineeringVo1．41No．18DOI：10．3969／j．issn．1001-3881．2013．18．023SegmentationAlgorithmofContourExtractionforImageEdgeBasedonFuzzyMorphologyHUANGWeiDepartmentofAutomation，ChongqingIndustryPolytechnicCollege，Chongqing401120，ChinaAbs

2、tract：Thereisextensiveapplicationofcontourextractionforimageedgeintheaspectofin-dustrialnondestructivetesting，assemblyandtestingofcomplexparts．Aimedatbeingeasytoproduceover-segmentationofremotesensingimageresultedintheimpactofnoiseetc，thepa-perexploredanalg

3、orithmforsegmentationofremotesensingimageincropsbasedonfuzzymorphology．Bymeansofopenandclosedoperationoffuzzymorphology，themethodfirstmakessmoothprocessingfortheimage，thenexecutethegradientcalculationbasedofmorpho-logicalgradientoperator，andfinallybasedoffu

4、zzymorphology，anditcarriesthoughtheseg—mentationforthegradientimageandachievestheimagetobesegmented．Thesimulationtooktheimagerecognitionasanexample，andtheexperimentdemonstratedthattheimagesegmen-tationalgorithmbasedonfuzzymorphologycanefectivelyeliminatethe

5、over·segmentationphe·-nomenon，distinguishtheimagecorrectly，andrealizetheseparationoftargetandbackground．Theresearchresultshowstheproposedmagesegmentationalgorithmisreasonableandfeasi—ble．Keywords：fuzzymorphology，gradientoperator，imagesegmentation，contourrec

6、ognitionagestructurefeaturebeingtakenintoaccountfor1．Introductionmathematicalmorphologyfully，ithasveryuniquesuperiority，anditswatershedtransformationisasortThetechnologyofimageedgecontoursegmenta—ofimagesegmentationmethodthatcanobtainaccu—tionhasveryimporta

7、ntpracticalvalueinimageanal-rateimageedge．Butitisverysensitivefornoiseinysisofremotesensingandvideoimageetcforlotsofimageprocessingandeasytoproduceover-segmenta--fieldssuchasindustrialproduction，targetdetectiontionphenomenon[4—8]．Aimedattheover-segmen-ofmil

8、itaryaffairs，agriculture，seasandoceans，pro-tationresultedinnoise，inordertoimprovetheimagetectionofenvironmentresourceandsoon[1—3]．segmentationquality，thepaperexploredasortofTheimagesegmentationisasorto