文献类型：期刊文献

中文题名：ERS SAR干涉测量技术用于区域尺度森林制图研究

英文题名：Regional Forest Mapping Using ERS SAR Interferometric Technology

作者：李增元[1] 庞勇[1] 陈尔学[1]

第一作者：李增元

机构：[1]中国林业科学研究院资源信息研究所

年份：2003

卷号：19

期号：4

起止页码：66-70

中文期刊名：地理与地理信息科学

外文期刊名：Geography and Geo-Information Science

收录：CSTPCD;;国家哲学社会科学学术期刊数据库;北大核心:【北大核心2000】；CSCD:【CSCD2011_2012】；CSSCI:【CSSCI2003】；

基金：国家863-308计划资助项目(863-308-14)

语种：中文

中文关键词：ILU影像;干涉测量土地利用影像;区域尺度森林制图;SAR;干涉测量技术;ERS;合成孔径雷达

外文关键词：ILU image; regional forest mapping

分类号：P285.23;P23

摘要：光学遥感数据由于受云雾的影响,很难在区域尺度上获得时间一致的图像,而微波遥感数据又一直无法提供足够的类型信息。雷达干涉测量技术的出现解决了这一瓶颈问题,雷达干涉相关值影像已经成功地用于土地利用分类。该文论述了干涉土地利用影像(ILU)的生成、基于ILU影像的森林分类方法、大区域ILU影像的几何纠正及镶嵌技术,并利用ERSSAR串行轨道数据生成的ILU影像成功地对我国东北三省进行了森林制图,其分类平均精度在82%以上。
This study aims to produce a baseline forest map of China using Interferometric Land Use (ILU) products generated from ERS-SAR Tandem data with the Interferometric Quick Look (IQL) system at ESRIN. Interferometric Land Use products exploit the potential of 1-day interferometric coherence for the retrieval of bio- and geo-physical parameters in land applications. An ILU image is a false colour composite consisting of the combined backscatter of two images and the difference and correlation between them: Red = interferometric coherence; Green = mean/minimum intensity; Blue = intensity difference. This particular RGB combination has been chosen to give land use types a true colour representation.The following contents are given in this paper: i) The algorithm of ILU image generating; ii) The decision tree classifier of forest and non-forest discrimination; iii) The georeference and mosaic method of long strip images. This paper demonstrates the feasibility and method for regional forest mapping using ILU image. A total accuracy of 82.4% has been achieved in Northeast China.