How does TerraVision represent terrain?
TerraVision uses a standard multiresolution pyramid representation for all of its terrain data. This is analogous to a quad tree structure. To do this, we chop up a large terrain model (imagery or elevation) into a grid of 128 x 128 pixel tiles. We also produce versions of the image over a range of resolutions (e.g. 2048×2048, 1024×1024, 512×512, 128×128), and chop each of these up into tiles. That way, we can directly access any part of the image at any desired resolution. This allows us to interactively browse terrain datasets over the Web by only streaming the parts of the images that we actually require for the current view. • What bandwidth does TerraVision require? TerraVision does not require any specific bandwidth in order to operate. TerraVision was designed so that the ability to navigate is independent of the speed at which data is received. You can always fly over the earth as fast as your computer can draw frames. Due to the multiresolution representation that TerraVision
