SURFACE APPROXIMATION OF COASTAL REGIONS: LR B-SPLINE FOR DETECTION OF DEFORMATION PATTERN

Abstract. Geospatial data acquisition of terrains produces huge, noisy and scattered point clouds. An efficient use the acquired requires structured compact representations. Working directly in a cloud is often not appealing. To face this challenge, approximation with tensor product B-spline surfaces attractive. It reduces description to relatively few coefficients compared volume original cloud. However, representation lacks ability adapt resolution shape local variations The result frequently that noise approximated have unwanted oscillations.Locally Refined (LR) were introduced challenge provide tool for approximating Geographic Information System In our LR based algorithm, iterative least-squares combined Multilevel Approximation reduce memory consumption. We apply approach sets from coastal regions Norway Netherlands, compare obtained raster method. further highlight potential volumes spatio-temporal visualisation deformation patterns.