一種基于小波變換和數(shù)據(jù)融合降噪的圖像邊緣檢測方法.pdf

1、Y766；31中山大學(xué)碩士學(xué)位論文一種基于小波變換和數(shù)據(jù)融合降噪的圖像邊緣檢測方法專業(yè)：計(jì)算數(shù)學(xué)學(xué)位申請人：羅婷導(dǎo)師及職稱：關(guān)履泰教授李小福副教授翥i簿名i委員：鱺甲“沙一“(1繡坼二零零五年四月AB盯RA(了ABSTRACTTheedgeofanimageisofIenconsideredasitsfIlnd鋤entalfeaturejnjmageprocessillg1thasbeenextens主Velyapp】jedinimag

2、epmccssjngandanalyzjngtecllnologjes，suchasfealuledescfjptjon，iIIlagere。09Ilition，imagescgmentation，imageenhancemencaIIdimagcdatacoInpressiononahigtlIeVeIEdgedetectionisabasicand沛portaIltsubjectiIIcomputerVisionandimagepr

3、ocessing，thequalityofwhichdetefm缸estlIere跚hofsubsequentProcessingThusitisane鑫ongoalfbrpeopletoquestanedgcdetectiona毽orith峨whichcansolVetheharmonyofedgcdetedionprecisionandre‘舡ainnoi9esIⅡthenoisedconditions，inofdertOiIIlp

4、roVethemethodpfoposcdbyWangeta一91，w“hwhichthefuseddalaafeobtainedbytheweightmeanvaIueofthedata丘omthcderivative叩erators，thispaperprcsentsanoiscrestr咖edmethodofimageedgedetectionbasedonwaVeleIlransfof腳anddataflIsionWi咖lhjs

5、memod，parfofthenojsescanbefemovedfirstBu￡thefearesomenoiSesremaiⅡedSincen地derivativc叩eratorsareseⅡs“iVetonoise，theyalsoenhancethe加iseswhentheydetectouttheedgesothedata宜omderiVativeoperatofsarefI】sedconlb證edwithrestrainiI

6、lgnoisesInhigh讎quency，thecoe茄cientsaretransactedbythresholdfinerandthenthe打nponantwavekt∞e蜘ciemsareselectedaccofdingby‘majorityprincipks’，subsequently，theweightsoffllsjoncanbeobtainedbyproportionofedgeillformatbnInlow丘eq

7、uency，theprinciplesofchoos協(xié)gcoe所cicntsarebasedolIedgcsofiⅡlagesFinally，thebin缸yedgeimageisproducedbythe～sed蛔agetran翰ctedbytllresholdoperationalldthinnedExperiIIlentalresults曲owlbattheedgccanbewelldetededoutand，allhesamel