Skip to Main Content
Application of a spline approximation method to computation and analysis of lidar-based digital elevation models is investigated to determine its accuracy and capability to create surfaces at different levels of detail. Quadtree segmentation that adapts to the spatial heterogeneity of data points makes the method feasible for large datasets. The results demonstrate the importance of smoothing for the surface accuracy and noise reduction. A tension parameter is effective for tuning the level of detail in the elevation surface. Simultaneous computation of topographic parameters is applied to extraction of sand dunes' features for assessment of dune migration and beach erosion.