Based On Sar Image Of Floods And Snow Disaster Area Extraction Algorithm Research

Each year, flood and snow disaster cause heavy economic loss and severe casualty in many areas. Synthetic Aperture Radar (SAR) with the ability of all-weather and high resolution has become an important means of natural disasters observation. The main contents in this thesis are water information extraction and snow cover retrieve based on SAR data.The common methods used in water information extraction at present are introduced and their application characteristics are analyzed. Then, the author used land-scatterometer to measure the microwave scattering characteristics of water and semi-flood soil, proposed using Otsu classification combing microwave scatter character to extract flood and semi-flood area. Also, the author used computer program to achieve the results of classification in areas near ShangRao city and GaoYou Lake, and analyzed the influence of local weather. After this, the comparing of multi-temporal classification by this method will provide different flood situations as many as eight.The thesis researched snow scattering model and compared with SAR data. Based on prior ways of snow cover retrieve, the author used multi-temporal change detection to delineate snow area and discussed the help of DEM data. Also, the author got the classification results in area near QingHai Lake and TianChang city through computer program. With comparing of the two results, the author found that, the effect of change detection way will be severely wreaked by people’s actives, the development of city and crop growth.At the last, the author analyzed using coherence to detect water body and snow cover area. Also he discussed its ability of disaster monitor. Then the author did an experiment to prove the usefulness in water classification.