3644

3443

كارشناسي ارشد

مرتعداري

اصفهان : دانشگاه صنعتي اصفهان ، دانشكده منابع طبيعي

1386

چهارده ، 85،[II] ص.: مصور ، نمودار ، جدول

ص.ع. به فارسي و انگليسي

جمال الدين خواجه الدين

عليرضا سفيا نيان

13/5/86

سعيد سلطاني ، محمد رضا وهابي

مهندسي منابع طبيعي

ID3443

به فارسي و انگليسي : قابل رويت در نسخه ديجيتال

Abstract The incresing number of sensor type for terrestrial remote sensing has necessitatedsupplementary efforts to evaluate and standardize data from the different aailablesensors Accurate and up to date global land cover data sets are necessary for variousglobal change research studies including climatic change biodiversity conservation ecosystem assessment and environmental modeling Therfore using of data numeralproduct of sensor with different resolution and specteral bands can different data forevalution and range management for users generate The aims of the present research was Comparison maps vegetation production of sensors IRS WiFS and MODIS on Semiromand Brojen regions Various preprocessing including image rectification was applied withgeo referencing of the image to a registered image with a RMSe 0 5 pixel for sensoreMODIS and RMSe 35 pixel for sensore IRS WiFS The atmospheric and topographiccorrections were applied using subtraction of dark objects method and the Lambert methodaccordingly Image processing including FCC PCA vegetation indices and supervisedclassification were employed to produce the land use and pasture production maps Fielddata collection were started on June 2005 on 800 000 ha and prolonged about 4 months Various vegetation types were sampled using stratified random sampling method 20random sampling points were selected and the percent cover was estimated Tow MODISdata sets and tow IRS WiFS data sets were used for this study Digital data and the indicesmaps were used as independent data and the field data as dependent variables The resultedmodels were processed on and the resulted images were categorized in 5 classes Finallythe produced maps were controlled for their accuracies Also post classification methodwas used to determine change detections Map land use of MODIS data sets produced with9 category and accuracy 79 61 and with Kappa 72 6 and map landuse ofIRS WiFSproduced with 8 category and accuracy 90 2 and with Kappa 77 9 Result showed mapsland cover produced with IRS WiFS data setes have very high accuracy land coverproducced with MODIS data setes

جمال الدين خواجه الدين

عليرضا سفيا نيان

سعيد سلطاني ، محمد رضا وهابي