遥感影像湖岸线提取方法研究

学号：***********密级：武汉大学本科毕业论文遥感影像湖岸线提取方法研究院(系)名称：电子信息学院专业名称：测控技术与仪器学生姓名：**指导教师：刘刚二〇一一年五月BACHELOR'SDEGREETHESISOFWUHANUNIVERSITYResearchofLakeShorelineExtractionBasedonRemoteSensingImageryCollege：SchoolofElectronicInformationSubject：TechnologyandEquipmentofObservingandControllingName：WeiYangDirectedby：GangLiuMay2011郑重声明本人呈交的学位论文，是在导师的指导下，独立进行研究工作所取得的成果，所有数据、图片资料真实可靠。尽我所知，除文中已经注明引用的内容外，本学位论文的研究成果不包含他人享有著作权的内容。对本论文所涉及的研究工作做出贡献的其他个人和集体，均已在文中以明确的方式标明。本学位论文的知识产权归属于培养单位。本人签名：_________________日期：_________________摘要湖泊的岸线是表征湖泊形态特征的一项重要内容,是划分湖泊与陆地管理区域的基准线。随着遥感对地观测技术的不断发展,卫星遥感影像在湖岸带地形图的更新中占据了愈来愈重要的地位。快速而准确地监测湖岸线的动态变化对于湖域的使用管理具有十分重要的意义。遥感技术具有快速、动态、大范围、宏观等突出优势。利用遥感图像更新湖岸带地形图的关键是湖岸线的提取,而湖岸线提取的实质是遥感图像边缘的提取。边缘中包含图像物体有价值的边界信息，这些信息可以用于图像理解和分析，并且通过边缘检测可以极大地降低后续图像分析和处理的数据量。图像理解和分析的第一步往往就是边缘检测，目前它已成为机器视觉研究领域最活跃的课题之一，在工程应用中占有十分重要的地位。本文首先对遥感影像进行灰度直方图统计计算,然后采用自动阈值法将水陆分离开，对图像进行中值滤波后分别采用Roberts算子、Sobel算子、Prewitt算子、LOG算子和Canny算子对遥感影像进行湖岸线提取，最后采用连通对象标注去除一部分的较小的边缘，得到较明显的遥感影像的湖岸线，最后对结果进行不同波段以及不同算法的比较和讨论。关键词：遥感；湖岸线；边缘检测；图像处理ABSTRACTLakeshorelineisoneofimportantfeaturesoflakemorphology,isthebasedirectrixdividestheregionoflakeandlandmanagement.WiththedevelopmentofremotesensingtechnologyforEarthobservation,satelliteremotesensingimagesplaysanincreasinglyimportantroleintheupdateofmapoflakeshorezone.Itisveryimportanttomonitorthedynamicchangeoflakeshorelinerapidlyandaccuratelyregardingthelakeareauseandmanagement.Remotesensingtechniquehasfast,dynamic,largescope,macroscopicandsuchprominentadvantages.Thekeytouseremotesensingimagestoupdatethemapoflakeshorezoneislakeshorelineextraction.Theessenceoflakeshorelineextractionisedgedetectionofremotesensingimage.Theedgeincludesvaluableinformationoftheimagewhichcanbeusedinimageunderstandingandanalysis.Andthroughedgedetection,wecangreatlyreducethecalculationofimageanalysisandprocessinginthefollowingstep.Usually,thefirststepofimageunderstandingandanalysisisedgedetection,andithasbeenthemostactivesubjectinthemachinevisionresearchfield,alsoitplaysanimportantroleinengineeringapplication.First,thisarticlecalculatesstathistogramtotheremotesensingimage，thenadopttheautomaticthresholdmethodtoseparatethelandandwater,thenextractslakeshorelineofremotesensingimagewithRobertsedge-detection,Sobeledge-detection,Prewittedge-detection,LOGedge-detectionandCannyedge-detectionafterusingmedianfiltertosmoothimages.Thenadoptandconnectthetargetandmarktoremovethesmalleredge,receivetheedgeofthelakestrandlineofmoreobviousremotesensing.Finally,theresultsofdifferentbandsanddifferentalgorithmsarecomparedanddiscussed.Keywords:RemoteSensing;LakeShoreline;EdgeDetection;ImageProcessing目录第1章绪论………………………..…………………………………………11.1遥感影像湖岸线提取的意义..................................................................11.2湖岸线探测技术......................................................................................11.2.1湖岸线现场探测技术....................................................................21.2.2遥感影像水边线提取技术............................................................21.3国内外研究现状......................................................................................21.3.1国外研究现状................................................................................31.3.2国内研究现状................................................................................31.4本文主要任务..........................................................................................3第2章水陆分离……………………………………………….……………42.1灰度直方图..............................................................................................42.2阈值分割..................................................................................................42.2.1全局阈值法....................................................................................52.2.2局部阈值法....................................................................................6第3章图像边缘提取方法…………….………………….………………83.1基于一阶导数法的边缘检测..................................................................93.1.1Roberts算子.................................................................................103.1.2Sobel算子....................................................................................103.1.3Prewitt算子...................................................................................113.2基于二阶导数法的边缘检测.................................................................113.2.1拉普拉斯算子..............................................................................123.2.2二阶方向导数..............................................................................143.3LOG算法...............................................................................................143.4Canny边缘检测算法............................................................................16第4章边缘提取仿真实验…………….……….…………………………194.1边缘提取实验概述................................................................................194.2图像阈值分割........................................................................................204.3图像平滑处理........................................................................................224.4边缘检测分割方法..................