【摘要】 "5.12"汶川8.0级地震后发生大量滑坡、泥石流,河流部分阻断,形成一系列堰塞湖,急需快速提取水体信息,为救援提供重要依据。本文以震后北川县城为例,利用LBV变换与NDVI(归一化植被指数)图像结合模型对研究区水体进行提取,对"福卫二号"影像进行LBV变换,确定B阈值,并提取NDVI指数,两者求交集得到水体影像。该模型理论依据充分可靠,计算机容易实现,具有较高的准确度,可以将水体与低密度覆盖的水植混合体区分开,提取结果符合实际情况,提取速度快,完全能满足灾害背景参数8h内提取的要求。更多还原

【Abstract】 In the Wenchuan MS 8.0 earthquake on May 12,2008,a large number of land-slides and mudslides occurred,which partially blocked rivers and formed dammed lakes.The information on water bodies is need to be fast extracted in order to provide important data for disaster relief.With Beichuan County as an example,water body information is extracted by combined using LBV transformation and NDVI(Normalized Differential Vegetation Index) images.By making the LBV transformation of FORMOSAT-2 images,determining B threshold and extracting NDVI index,the water body images are obtained and an extraction model is set up.The model is on reliable theoretical basis and easy to implemented by computer,can discriminate water bodies from low-density water-plants mixture and gives the results in consisten with the actual situation.The extraction speed is so high that it meets the requirements of extracting disaster background parameters within 8 hours.更多还原