Level-of-detail-based rendering and compression for 3D GIS

Since geographic data sets tend to be quite large, problems arise when visualizing large GIS data. Terrain visualization is also a difficult problem for 3D GIS applications requiring accurate images of large datasets at high frame rates. In this paper we present algorithms for compressing and visualizing large 3D GIS datasets. We propose two algorithms which have different purposes. The goal of the first algorithm is to minimize the number of bits needed to encode connectivity information of TIN (triangulated irregular network) by using Delaunay triangulation. We found that most of the terrain triangulations are quite similar to the plane-projected Delaunay angular mesh. This experimental work shows that more than 93% of the edges of a common terrain data are included in Delaunay triangulation. By exploiting this result, we can on average compress the common terrain data by 0.23 bits per vertex. The second algorithm is an interactive visualization algorithm for 3D GIS datasets using surface simplification. Whenever a viewer gets closer to the 3D GIS object, a geometry with a high level of detail is visualized. Our algorithms achieve compression ratios that are better than those of the currently available algorithms, and moreover, it allows progressive and interactive visualization or the 3D GIS datasets.