understanding sparse point cloud and dense point cloud

[mpatalberta]

In agisoft you:
1- Load you picture
2- align picture ( result is a sparse point cloud
3- then generate a dense point cloud
Does agisoft use the sparse point cloud from align to generate the dense point cloud?
If I filter the sparse point cloud does affect the dense point cloud?
How does agisoft go from a few points in the sparse point cloud to the dense cloud ? Does the software interpolate the data?
Thanks,
Pat,

[Alexey Pasumansky]

Hello Patrick,

Sparse cloud is used to find out which image pairs are overlapping (have sufficient number of common valid tie points), then during the dense cloud generation the depth maps are created for such overlapping pairs and merged together.

I think that you may want to optimize the camera alignment after removing some points from the tie point cloud - thus interior and exterior orientation parameters for the cameras will be refined. Improved quality of the camera alignment should also result in better quality of the depth maps.