DEM multi-scale representation based on wavelet multiresolution analysis

In the same area, the terrain details described by DEM are gradually generalized and primary terrain features are only retained in the coarser-resolution DEM. In this paper, wavelet multiresolution analysis and Radical Law selection principles were combined to model the terrain generalizing processes and derive three different scale-parameter DEMs based on DEM data generated by large-scale digital topographic map. Comparison among the contour lines and elevations of different scale-parameter DEM shows smaller scale-parameter DEMs retained the features of landform morphology such as terrain outline, ridge and gully trend of large-scale DEM. Furthermore, the results show that the terrain surface is gradually smoothing and the derived DEMs still have accurate spatial structure, by analyzing the change of slope gradient, profile curvature and stream network extracted from different scale-parameter DEMs. It could model arbitrary-scale DEM using wavelet multiresolution analysis and the Radical Law.