简介:这篇论文为遥远地察觉到的数据,结果的分类的模糊的度被作为由一个常规算法与那作比较减少聚类的fuzzyc工具描述一个改进算法:也就是说,分类精确性被增加。这被在单个班的水平合并协变性矩阵而非假定全球的完成。从一个爱丁堡郊外的陆地封面的一个模糊分类的实验结果为聚类的fuzzyc工具证实了新算法的改进性能,特别地当模糊也在假定参考书数据被提供时。
简介:一九九四年·测绘法规·山西省人民代表大会常务委员会公告(1-4)…………山西省测绘管理条例(1-5)……………………………关于《山西省测绘管理条例(草案)》的说明……………王银焕(1-9)…………………………………………关于《山西省测绘管理条例(草案)》修改情况的说明…李树森(1-12)………………………………………关于《山西省测绘管理条例(草案)》(修改稿)修改情况的说明李红星(1-14)………………………………………贯彻《山西省测绘管理条例》,促进山西测绘事业的发展夏志朴(1-16)………………………………………·重点工程测量·关于万家寨引黄工程测量控制网的几点意见…………何万龙(1-19
简介:
简介:ItiswellknownthatLandsatTMimagesarethemostwidelyusedremotesensingdatainvariousfields.Usually,ithas7differentelectromagneticspectrumbands,amongwhichthesixthonehasmuchlowergroundresolutioncomparedwiththeothersixbands.Nevertheless,itisusefulinthestudyofrockspectrumreflection,geo-thermalresourcesexploration,etc.ToimprovethegroundresolutionofTM6tothelevelasthatoftheothersixbandsisaproblem.Thispaperpresentsanalgorithmbasedonthecombinationofmulti-variateregressionmodelwithsemi-variogramfunctionwhichcanimprovethegroundresolutionofTM6by"fusing"thedataofothersixbands.Itincludesthefollowingmainsteps:(1)testingthecorrelationbetweenTM6andoneofTM1-5,7.IfthecorrelationcoefficientbetweenTM6andanotheroneisgreaterthanagiventhresholdvalue,thenselectthebandtotheregressionanalysisasanargument.(2)calculatingthesizeofthetemplatewindowwithinwhichsomeparametersneededbytheregressionmodelwillbecalculated;(3)replacingtheoriginalpixelvaluesofTM6bythoseobtainedbyregressionanalysis;(4)usingimageentropyasameasurementtoevaluatethequalityofthefusedimageofTM6.ThebasicmechanismofthealgorithmisdiscussedandtheVC++programforimplementingthisalgorithmisalsopresented.Asimpleapplicationexampleisgiveninthelastpartofthispaper,showingtheeffectivenessofthealgorithm.