中文摘要：

采用多层紧支撑径向基函数对离散数据内插生成DEM.首先对采样数据进行预处理,在原始数据上删除最小距离点对上的点,每次删除后将剩下的数据点存储并作为下一次删除操作的数据源,重复删除操作至点数为给定的初值;然后在已存储的数据集中选取均匀度较大的m个子集作为插值的数据源,分别进行紧支撑径向基函数插值,将上一层的误差作为下一层的插值属性值,求解插值矩阵系数.将规则格网点作为插值点通过插值矩阵内插生成DEM,并且与各种径向基函数插值方法进行比较.实验分析表明,基于多层紧支撑径向基函数的DEM具有较高的插值精度,可作为一种DEM插值方法.

英文摘要：

Scattered data is interpolated to generate DEM based on multistep compactly supported radial basis function. Firstly, the original scattered sampling data is preprocessed. The point with minimum distance in original data is deleted and the rest da- ta is stored as the source of following deletion operation. Repeating the deletion operation until the number of point equals to the given value. Then,m subsets with higher evenness are seleeted as the interpolation data source and are interpolated with com- pactly supported radial basis function. In the interpolation process, take the errors of the previous layer as interpolation attribute values of the next layer, in this way the interpolation coefficient matrix is solved. Taking regular square grid points as interpola- tion points, based on which DEM is generated. In this paper, different kinds of radial basis function interpolation methods were compared and analyzed. The result shows that the reconstruction of DEM based on multistep compactly supported radial basis function has higher accuracy and can be used as an effective method of DEMrs reconstruction.