近红外波段ETM^+影像的积雪提取方法  

Snow cover extraction method based on the near-infrared band of ETM^+ image

作　　者：姜萍 王晓威 

机构地区：[1]新疆气象服务中心,乌鲁木齐830002 [2]中国地质大学(武汉)信息工程学院,武汉430074

出　　处：《测绘科学》2017年第11期41-46,共6页Science of Surveying and Mapping

摘　　要：针对提高积雪信息提取精度的要求,为了消除积雪覆盖的结冰水体、薄雪覆盖区以及山体阴影等对于积雪提取的影响,以Landsat-7ETM+为数据源,对近红外波段在积雪信息提取中的优越性进行了探索,提出了一种基于近红外波段和归一化差分积雪指数的积雪提取方法。对典型实验区进行了对比实验分析,结果表明,本文算法能有效减少在结冰水体、薄雪覆盖区以及山体阴影等区域的漏分、误分像元数,获得比SNOMAP算法更佳的积雪识别效果,提高积雪提取的准确性。In order to extract snow cover information high-precisely and to eliminate the impacts of the frozen river water,thin snow area and the shadows of mountain on the extraction,in this paper,by taking Landsat-7 ETM~＋ data as resources,the superiority of near-infrared band to abstract snow fraction was explored,and a method of snow cover extraction was proposed by integrating the near-infrared band and NDSI.The contrast experiments are completed in two representative zones.It is found that the proposed extraction method can effectively reduce the pixel numbers of unclassified and misclassified,so as to gain a better effect of snow detection than SNOMAP algorithm,and to improve the accurate of extraction result.

关 键 词：积雪提取 近红外波段 归一化差分积雪指数 SNOMAP算法 

分 类 号：P237[天文地球—摄影测量与遥感]