摘要
MODIS遥感数据时间分辨率较高,将其应用于河道监测优势明显。但是对于山区多云地区,MODIS数据时间分辨率高的优势无从发挥,且山区沟壑众多,河道狭窄,水体像元多为混合像元,利用现有方法提取水体遥感影像难度较大。为了能够将MODIS数据的优势运用于山区河道监测,选定大渡河流域作为典型研究区域,在对MODIS数据中的典型地物进行采样分析后,首先采用基于小波变换改进的同态滤波削弱云雾,然后对归一化植被指数法(RVI)进行改进,得到了薄云覆盖下山区水体指数(CMWI),消除了薄云的干扰,实现了山区沟壑的模糊处理,突出了水体。图像增强后,河道水体清晰可辨,适用于山区河道遥感监测工作。
Using MODIS remote sensing data to monitor rivers brings obvious advantages since it has high temporal resolution.However,there is too much cloud in the mountain area of western Sichuan Province to show the advantages of MODIS data,also many gullies exist in there and the river is very narrow.It′s too hard to extract water-body with existing method since the water-body pixels are mixed pixels.In order to apply the advantages of MODIS data to monitor river,we choose Dadu River watershed as the typical research area.After sample analysis of typical surface features of MODIS data,firstly apply homomorphic filtering based wavelet transform to weaken cloud effect,and then to get water-body index of mountain area by improving RVI method,therefore the disturbance of cloud has been eliminated.In this way,the gully imagine are fuzzily treated and the water-body imagine is enhanced and easily recognized.The processed image can be applied to monitor rivers in mountain area.
出处
《人民长江》
北大核心
2010年第21期14-17,共4页
Yangtze River
关键词
MODIS
山区河道
水体提取
河道遥感监测
去云
MODIS
river in mountain area
extraction of river water-body imagine
river channel monitoring by remote sensing
cloud-removing treatment
