資源描述：
《邊緣檢測(cè)微分算子的分析及在醫(yī)學(xué)圖像中的應(yīng)用》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在工程資料-天天文庫。
1、Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/7972542ThefrequencyfeaturesandapplicationofedgedetectiondifferentialoperatorsinmedicalimageArticleinShengwuyixuegongchengxuezazhi=Journalofbiomedicalengineering=Shengwuyixuegongchengxuezazhi·Mar
2、ch2005Source:PubMedCITATIONSREADS4295authors,including:JianWuHuiDingTsinghuaUniversityTsinghuaUniversity42PUBLICATIONS257CITATIONS69PUBLICATIONS178CITATIONSSEEPROFILESEEPROFILEGuangzhiWangTsinghuaUniversity246PUBLICATIONS714CITATIONSSEEPROFILESomeoftheauthorsofthispublicationarealsoworkingonth
3、eserelatedprojects:MedicalRoboticsforInterventionalSurgeryViewprojectRehabilitationEngineeringViewprojectAllcontentfollowingthispagewasuploadedbyGuangzhiWangon04January2017.Theuserhasrequestedenhancementofthedownloadedfile.生物醫(yī)學(xué)工程學(xué)雜志2005;22(1)∶82～85JBiomedEng邊緣檢測(cè)微分算子的分析及在醫(yī)學(xué)圖像中的應(yīng)用$山吳 劍 丁 輝 王廣志 丁
4、海曙 周伊豆頁(清華大學(xué)醫(yī)學(xué)院生物醫(yī)學(xué)工程系,100084) 摘要 邊緣檢測(cè)是醫(yī)學(xué)圖像處理必不可少的一步,采用微分算子檢測(cè)邊緣是最常用也是處理效果比較好的一種方法。我們從頻域角度出發(fā),分析了Roberts算子、Prewitt算子、Sobel算子以及拉普拉斯算子的頻譜特性。提出了在使用和構(gòu)造微分算子時(shí),要充分考慮它們的頻譜特性。不同的微分算子對(duì)不同邊緣檢測(cè)的敏感程度是不同的,因此對(duì)不同類型的邊緣提取,應(yīng)該采用對(duì)此類型邊緣敏感的算子進(jìn)行邊緣提取。最后通過實(shí)際的MRI圖像處理驗(yàn)證了邊緣檢測(cè)算子選擇的重要性。關(guān)鍵詞 邊緣檢測(cè) 微分算子 醫(yī)學(xué)圖像TheFrequencyFeaturesan
5、dApplicationofEdgeDetectionDifferentialOperatorsinMedicalImage$WuJianDingHuiWangGuangzhiDingHaishuZhouYiyi(DepartmentofBiomedicalEngineering,SchoolofMedicine,TsinghuaUniversity,Beijing100084,China)AbstractEdgedetectionisanabsolutelynecessarystepinmedicalimageprocessing,andtheuseofdifferentialo
6、peratorstodetectedgeisoneofthemostcommonandeffectivemethods.InthispaperareanalyzedthefrequencyfeaturesoftheRobertsoperator,Prewittoperator,SobeloperatorandLaplacianoperatorfromtheviewpointoffrequencydomain,anditisproposedthatthefrequencyfeaturesofthedifferentialoperatorsshouldbeconsideredwhend
7、ifferentialoperatorisbeingusedand?orconstructed.Becauseedgedetectionoperatorissensitivetotheedgetype,theappropriateoperatorshouldbeadoptedindifferentedgetypedetection.Finally,theimportanceandnecessityofselectingedgedetectionoperatorarev
