摘要
利用HJ-1星2009年"莫拉克"台风前后获取的2个时相图像,通过去相关拉伸(Decorrelation Stretch)、光谱信息增强和最大似然法分类提取滑坡、泥石流区域,并结合TRMM(Tropical Rainfall Measuring Mission)卫星降雨量数据和DEM(Digital Elevation Model)资料,分析该台风降水导致的滑坡、泥石流情况。结果表明,HJ-1星能有效地监测滑坡、泥石流灾害。基于去相关拉伸光谱信息增强的最大似然法分类能得到较好的分类结果。"莫拉克"台风带来的暴雨所引起的山体滑坡和泥石流导致植被和居民点变为裸地,分类图和3D模拟图上都体现了这种变化。
In this paper, two HJ - 1 images before and after typhoon Morakot were used. Decorrelation stretch （DS） and maximum likelihood classification （MLC） were adopted to extract the regions of landslide and debris flow. Furthermore, TRMM precipitation data and digital elevation model （DEM） were combined to analyze landslide and debris flow. The results show that landslide and debris flow can be monitored effectively by HJ - 1. Good classification results are obtained： the overall accuracies of two phases are above 90%, and Kappa coefficients are higher than 0.9. The rainstorm brought by typhoon Morakot led to landslide and debris flow, so vegetation and settlements were changed into bare land, which is reflected clearly on the classification map and 3 D map.
出处
《国土资源遥感》
CSCD
2011年第1期106-109,共4页
Remote Sensing for Land & Resources
基金
国家"863"课题(编号:2006AA120101)
国家自然科学基金项目(编号:40871158/D0106)
浙江省科技计划项目(编号:2007C22028)共同资助
