关键词: DEM, 稀疏点云, 融合, 高斯核函数, 加权迭代

Abstract: There will be sparse ground points that point cloud datas produced by airborne LiDAR in some local areas. When we use them to construct DEM, the problem of "triangle patch" will occur, which seriously affects the quality of DEM. This paper proposes a fusion method of sparse ground point cloud datas and existing DEM. The sparse point cloud as a high-precision control points, on the premise of keeping the topographic features of the original DEM as much as possible, the local elevation correction of DEM is carried out through Gaussian kernel function weighted iterative interpolation algorithm to realize the consistent fusion of sparse point cloud and DEM. The experiment shows that the fused point cloud datas are well supplemented, the new DEM has a natural topography. At the sametime,the accuracy is improved to a certain extent after fusion, its reliability is significantly improved in weak precision areas.

Key words: DEM, sparse ground points, fusion, Gaussian kernel function, weighted iteration