基于对象信息的遥感影像分类研究

基于对象信息的遥感影像分类研究

ID:32240656

大小:5.41 MB

页数:102页

时间:2019-02-02

基于对象信息的遥感影像分类研究_第1页
基于对象信息的遥感影像分类研究_第2页
基于对象信息的遥感影像分类研究_第3页
基于对象信息的遥感影像分类研究_第4页
基于对象信息的遥感影像分类研究_第5页
资源描述:

《基于对象信息的遥感影像分类研究》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、ABSTRACTABSTRACTComparedwiththelowormiddleresolutionimage,thehighresolutionRemoteSensingimagehasricherstructureandtextureinformation.Theresultoftraditionalstatisticalclassificationtechnologybasedonpixels’spectrumcontaininglotsof"PepperandSalt"noiseshasav

2、erylowaccuracy.Andthephenomenaof"thesamethingwithdifferentspectrums"and"differentthingswiththesamespectrum"cannotbedistinguished.Thosearealltheinevitablelimitationsofthetraditionalstatisticalclassificationalgorithmbasedonpixels’spectrum.Theideaofobject-o

3、rientedisintroducedintotheinformationextractiontechnologyfromthehighresolutionimageinthisthesis.Thistechnologyproduceshomogeneousimageobjectsthroughsegmentationtechnology,andprovidesawaytoanalyzeobject’sfeatures.Atlastitcarriesouttheinformationextraction

4、byfuzzyclassification.Analgorithmofimagesegmentation,basedoncolorandspatialinformation,isproposedinthisthesis.Theinitialregionsarehierarchicallymergedbasedontheregiondistancedefinedbycolorandspatialinformation.Andacriterionisproposedtodecidetheterminatio

5、nofthemergingprocess.Finally,theaccuracyofthesegmentationresultisashighastheresultmadebyeCogniton,“theNo.1EnterpriseImageIntelligenceProcessingSoftware”.Whenextractingtheobject’sfeaturesinthisthesis,wenotonlyconsiderthespectraldifferencebetweentheregions

6、,andshapemeasures,texturemeasures,positionmeasuresaswell.Onceusingtheseaddedobject’smeasures,wecoulddistinguishdifferentcoversonthegroundwithsamespectral.TakingatypicalregionofTongzhouDistrict,BeijingCityasthetestarea,acasestudyonSPOT-5imageclassificatio

7、nwithobject-orientedapproachiscarriedoutattheendofthisthesis.Inordertoverifytheaccuracyofobject-orientedclassification,acomparisonbetweenthisapproachandclassicalclassificationapproacheshasbeencarriedout.Thecasestudyshowsthattheapplicationofobject-oriente

8、dapproachonhighresolutionsatelliteimageclassificationnotonlycanimprovetheoverallclassificationaccuracy,butalsohasmoresemanticinformation,reducesthe“PepperandSalt”noises,andovercomessuchphenomenonasdifferentthingswiththesam

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。