遥感影像去除云阴影的高山区冰湖信息提取方法

A Cloud Shadow Removal Method for the Extraction of High Mountain Lake

利用光学影像提取高山区的冰湖信息时,云阴影和地形阴影是影响冰湖信息提取精度的重要误差源。为减少这一影响,本文提出并实验了一种利用水体指数法提取冰湖信息,利用光谱夹角阈值法、DEM生成的坡度图及地形晕渲图分别消除云阴影、地形阴影的综合方法。将论文算法在两幅存在云阴影和地形阴影的TM影像上进行了实验验证,实验结果表明,本文的方法能够有效消除云阴影和地形阴影对冰湖信息的影响,提高了光学影像的利用率,改善了冰湖信息提取的精度。

Many optical images are always covered by clouds,and the shadows produced by clouds and terrain are significant sources of noise of mapping glacial lakes in high mountain area.This paper developed a new method,which applies the normalized difference water index(NDWI)to delineate the glacial lakes,and uses the spectral angle method to eliminate the cloud shadow,and uses the slope maps and the shaded relief maps derived from DEM.The present method is applied on two TM images,which include cloud shadows and terrain shadows respectively.The result indicates the developed method can eliminate cloud shadows and terrain shadows effectively for the glacial lake delineation.In addition,this method can make more optical images to be used and improve the accuracy of glacial lake detraction.