Digital Elevation Models from Contour Lines

♦ Sampling density along the contours Interpolation of the irregular set of data points to a regular grid Valid theorems for sampling density have only recently been developed. Interpolation methods have to deal with the 'flat triangle' problem, as well as preservation of the accuracy of slope. Meaningful slopes are increasingly important in applications such as soil erosion and run-off modelling. The 'flat triangle' problem appears when the vertices of triangles have the same elevation which, for example, occurs at ridges and valleys. The problem is solved by generating a medial axis transform, or skeleton, between the curves. To arrive at DEMs with accurate slopes in which the 'flat triangle' problem is solved, we propose the following approach: Generate skeleton points and ignore skeletons between contours Assign elevations to these skeleton points Eliminate 'flat triangles' bythe insertion of these skeleton points into the original dataset Estimate slope information at each data point Perform weighted-average interpolation using the estimated slope information The approach is described and examined below, with the assumption that the data is sufficiently well sampled.