一、实验目的ENVI是一套功能齐全的遥感图像处理系统,是处理、分析并显示多光谱数据、高光谱数据和雷达数据的高级工具。此次实习主要是学习一些关于ENVI的基本操作,如:图像预处理,影像分析,图像增强,几何校正,监督分类以及专题制图等步骤。二、实验数据LE71440292000268SGS00.tar.gzELEVATION_SOURCE=GLS2000PROCESSING_SOFTWARE=LPGS_9.1EPHEMERIS_TYPE=DEFINITIVESPACECRAFT_ID=Landsat7SENSOR_ID=ETM+SENSOR_MODE=SAMACQUISITION_DATE=2000-09-24WRS_PATH=144BAND_COMBINATION=123456678PRODUCT_UL_CORNER_LAT=45.5786828PRODUCT_UL_CORNER_LON=84.0750064PRODUCT_UR_CORNER_LAT=45.6157964PRODUCT_UR_CORNER_LON=87.2821725PRODUCT_LL_CORNER_LAT=43.5718357PRODUCT_LL_CORNER_LON=84.1739972PRODUCT_LR_CORNER_LAT=43.6064525PRODUCT_LR_CORNER_LON=87.2726073PRODUCT_UL_CORNER_MAPX=271800.000PRODUCT_UL_CORNER_MAPY=5051400.000PRODUCT_UR_CORNER_MAPX=522000.000PRODUCT_UR_CORNER_MAPY=5051400.000PRODUCT_LL_CORNER_MAPX=271800.000PRODUCT_LL_CORNER_MAPY=4828200.000PRODUCT_LR_CORNER_MAPX=522000.000PRODUCT_LR_CORNER_MAPY=4828200.000三、实验内容1.遥感数据下载2.数据格式转化及多波段的合成3.图像边框的裁剪4.图像的镶嵌6.图像几何校正7.裁剪8.监督分类9.专题出图四、实验步骤1.遥感数据下载输入地名及日期,搜索符合自己通途的数据,查看并下载2.数据格式转化及多波段的合成由于下载图像的格式tif格式,需要转成img格式。并将单波段融合成为多波段,这一步在erdas中完成。tif转img单波段融合成多波段3.图像边框的裁剪通过主图像窗口-tools-regionofinterest-roitool,然后自己在图像上定义出一块感兴趣区(实际中难于确定严格准确的裁剪界限)。4.图像的镶嵌Map→mosaicking→georeferenced启动镶嵌模板6.图像几何校正设置参数,参数在原始数据中的txt中可以查看启动校正模块,并利用google地球进行校正7.裁剪分别打开需要裁剪的图像和所在区域的shapefile文件将矢量文件转成ROI利用ROI裁剪图像选择主菜单→basictools→subsetdataviaROIS,选择裁剪图像在spatialsubsetviaROIparameters中,选择由矢量生成的ROI,在“MaskpixelsoutsideofROI?”项中选择yes,MaskBackgroundValue为08.监督分类打开裁剪后的图像,以RGB:543显示,在display中选择Overlay→Regionofinterest样本选择完成后,在Option→ComputeROISeparability,选择分类图像,查看分离性InputFile:裁剪ROIName:(Jeffries-Matusita,TransformedDivergence)水体[Blue]36084points:农田[Green]18122points:(2.000000002.00000000)沙漠[Sienna]200333points:(1.999731782.00000000)未利用土地[Orchid]57805points:(1.999994942.00000000)山体[Purple]355228points:(1.982966922.00000000)农田[Green]18122points:水体[Blue]36084points:(2.000000002.00000000)沙漠[Sienna]200333points:(1.999548201.99999924)未利用土地[Orchid]57805points:(2.000000002.00000000)山体[Purple]355228points:(1.987447101.99928798)沙漠[Sienna]200333points:水体[Blue]36084points:(1.999731782.00000000)农田[Green]18122points:(1.999548201.99999924)未利用土地[Orchid]57805points:(1.987001772.00000000)山体[Purple]355228points:(1.846355611.96524858)未利用土地[Orchid]57805points:水体[Blue]36084points:(1.999994942.00000000)农田[Green]18122points:(2.000000002.00000000)沙漠[Sienna]200333points:(1.987001772.00000000)山体[Purple]355228points:(1.999873192.00000000)山体[Purple]355228points:水体[Blue]36084points:(1.982966922.00000000)农田[Green]18122points:(1.987447101.99928798)沙漠[Sienna]200333points:(1.846355611.96524858)未利用土地[Orchid]57805points:(1.999873192.00000000)PairSeparation(leasttomost);沙漠[Sienna]200333pointsand山体[Purple]355228points-1.84635561水体[Blue]36084pointsand山体[Purple]355228points-1.98296692沙漠[Sienna]200333pointsand未利用土地[Orchid]57805points-1.98700177农田[Green]18122pointsand山体[Purple]355228points-1.98744710农田[Green]18122pointsand沙漠[Sienna]200333points-1.99954820水体[Blue]36084pointsand沙漠[Sienna]200333points-1.99973178未利用土地[Orchid]57805pointsand山体[Purple]355228points-1.99987319水体[Blue]36084pointsand未利用土地[Orchid]57805points-1.99999494水体[Blue]36084pointsand农田[Green]18122points-2.00000000农田[Green]18122pointsand未利用土地[Orchid]57805points-2.00000000分类统计结果分类精度评价混淆矩阵ConfusionMatrix:H:\123456\jianduOverallAccuracy=(646026/667344)96.8055%KappaCoefficient=0.9486GroundTruth(Pixels)Class水体农田沙漠未利用土地山体Unclassified00000水体[Blue]335922000286农田[Green]751790748019沙漠[Sienna]346119581353615580未利用土地[O402166572690山体[Purple]4915423060339115Total360841812220033357805355000GroundTruth(Pixels)ClassTotalUnclassified0水体[Blue]336208农田[Green]18049沙漠[Sienna]212024未利用土地[O59439山体[Purple]341624Total667344GroundTruth(Percent)Class水体农田沙漠未利用土地山体Unclassified0.000.000.000.000.00水体[Blue]399.550.000.000.000.08农田[Green]0.2198.810.020.000.01沙漠[Sienna]0.090.3497.740.934.39未利用土地[O0.010.001.0899.070.00山体[Purple]0.140.851.150.0095.53Total100.00100.00100.00100.00100.00GroundTruth(Percent)ClassTotalUnclassified0.00水体[Blue]35.43农田[Green]2.70沙漠[Sienna]31.77未利用土地[O8.91山体[Purple]51.19Total100.00ClassCommissionOmissionCommissionOmission(Percent)(Percent)(Pixels)(Pixels)水体[Blue]30.790.45286/36208162/36084农田[Green]0.791.19142/18049215/18122沙漠[Sienna]7.652.2616211/2120244520/200333未利用土地[O3.650.932170/59439536/57805山体[Purple]0.734.472509/34162415885/355000ClassProd.Acc.UserAcc.Prod.Acc.UserAcc.(Percent)(Percent)(Pixels)(Pixels)水体[Blue]399.5599.2135922/3608435922/36208农田[Green]98.8199.2117907/1812217907/18049沙漠[Sienna]97.7492.35195813/200333195813/212024未利用土地[O99.0796.3557269/5780557269/59439山体[Purple]95.5399.27339115/355000339115/3416249.专题出图1、虚拟边框设置1)在主显示窗口菜单栏中选择File→Preferences,打开DisplayParameters对话框,设置虚拟边框的边界值和颜色。2)点击OK完成虚拟边框的设置。2、公里网设置ENVI支持同时显示像素公里网、地图坐标公里网以及地理坐标(纬度/经度)网。添加或修改地图影像公里网:1)主显示窗口菜单栏中选择Overlay→GridLines。打开GridLineParameters对话框中,显示默认的公里网设置。2)设置公里网属性参数。选择Options→EditMapGridAttributes或EditGeographicAttributes获EditPixelAttributes修改所选公里网的属性3)点击OK,完成参数设置。4)在GridLineParameters对话框中,点击Apply将新的公里网应用到影像图中。3、注记要素操作1)在主显示窗口菜单栏中选择Overlay→Annotation,打开Annotation对话框。2)从A