I recently mapped a large area, and due to circumstances, I had to fly it at 200 feet. I used the smart oblique technique with a 75/75 overlap. Not ideal, but we were running out of daylight, so I had to shorten the mission time.
As the requested deliverable is a model, I decided to use Reality Scan 2.0.1. When I reconstruct the model at High detail and zoom in, it becomes transparent, as if there aren’t enough points. At the normal detail setting, it is fine. The model remains solid when zooming in. I used ChatGPT to understand the impact of the different parameters in the reconstruction settings. I reduced the Detail decimation factor from 1 to 0.8.
Here are several screenshots showing this. I also show my settings. Any idea why this is happening and how I could improve it when using high detail settings?
This is just a GPU limitation of your computer. The meshes bigger as this limitation are shown as point cloud and not mesh. To check the setting go to WORKFLOW/Application/Settings/Rendering settings This will show you how big mesh are you able to show as solid.
Thank you for your reply. I increased it incrementally to 250M and saw an improvement. There is also the max triangles to display, which I set to 142M (80%).
What I don’t understand is that when I export the model in OBJ format, a friend who opens it sees the same in Blender. So does the render setting affect the export?
How many triangles has your model? For your GPU you’ll see the solid mesh only for 142 M tris.
The render settings doesn’t affect the exported mesh. I suppose it is too big to be shown in Blender or if it is georeferenced, then there could be issues with big distance from the zero coordinates.
