用遗传算法, 从多角度后向散射观测值同时反演陆地表面粗糙度(小尺度起伏方差与相关长度、大尺度起伏坡度)和土壤体湿度的多维特征性参数。粗糙陆地表面散射用Kirchhoff近似的稳相法与微扰法相结合的双尺度模型计算, 以构造代价函数。遗传算法反演结果与车载多角度遥感时的陆地实测值作了比较, 取得了十分良好的结果。为遗传算法在遥感反演中的新应用提供了实例。

A novel genetic algorithm is developed for multi parameter retrieval of land surface roughness and soil moisture. The theoretical model of two scale rough surface is employed for computation of the cost function. Retrieved results by this genetic algorithm are well compared with ground truth data. This study presents a new example of the genetic algorithm for multi parameter retrieval in remote sensing.