Note: Colored cones were used as placeholders for actual vegetation geometry objects. No billboards or imposters were used and full geometry was represented (40 triangles per cone). At this moment in time, we were running around 15 frames per second for 250k trees in the scene. We had targetting 1M vegetation objects at 30fps when fully LOD'd vegetation objects are used.
Example of TIGER road data and vegetation placement.
Where the satellite imagery runs out, we use a color ramp
based on elevation values to fill in the gaps.
Vegetation aids the eye in seeing the true shape of
the terrain (e.g. hills, ridgelines).
Terrain surface here is the national land cover dataset.
Vegetation on satellite imagery.
More satellite imagery with vegetation and roads.
Mix of satellite and psuedo colored terrain.
Hard to determine distance/range without terrain features.
Now we have a clearer idea of landscape's scale.
How close are we to the ground?
Wow! A lot closer than we thought!
Trees have arrived!
Trees have arrived!
Trees have arrived!