基于云雾与晴空地表混合像元的云雾检测算法

An Algorithm for Separating Cloud-Fog from Surface Features Based on Mixed Pixels

从专家目视判别晴空地表与云雾的原理出发,针对不同时区晴空地表与云雾分离检测算法的难点进行了研究,提出一种适应于不同时区的白天晴空地表与云雾分离的遥感检测算法。应用10幅不同时区MODIS图像对算法的可行性、稳定性及精度进行评定的结果表明,该算法可行性较强,稳定性较高,且不需人工干预。

The difficulties in separating cloud - fog from surface features in different time zones were studied in this paper. Based on the principle of visual interpretation, the authors have developed an adaptive threshold algorithm for separating cloud -fog from surface features. The parameters were determined based on MODIS data, and the stability, feasibility and precision of this algorithm were verified based on validation data. According to the results, this algorithm has strong stability and feasibility and can be performed automatically.