Do I need to reoptimize after adding a surface to the point cloud?

I’m generating an orthomosaic map from about 550 images over 5 acres. This area has several buildings on it which rendered the roof edges poorly in the initial ortho.

Based on the Pix4D help articles, I went back in and added surfaces to the roofs that were having problems. Of course, adding surfaces adds manual tie points. From what I’ve read, you should always reoptimize when creating MTPs. I’ve done this when joining projects before and I do re-optimize before that.

Do you need to reoptimize after adding surfaces (which would trigger then need to reprocess step 2)? If I understand correctly then, any project that requires surfaces to be added would need step 2 processed twice (once initially to get the point cloud created which is used to create the surfaces, then once after adding the surfaces)? My problem is that I already spent a lot of time cleaning up the point cloud before I realized I needed surfaces so if I reoptimize, that’s going to delete step 2 output, which means I need to clean up my point cloud again, right?

If I’m correct in all of this, would a proper workflow look like this:

  • Process steps 1 & 2
  • Add surfaces to areas you think might need them
  • Reoptimize
  • Reprocess step 2
  • Clean point cloud (ensuring no further surfaces are needed)
  • Process step 3

Hi @scott4 ,

Despite it is true that, for each vertex, a Manual Tie Point is created, it is possible but not required to reoptimize the project. More information at How to draw a surface in the rayCloud.

It is enough to draw a surface to have an impact on the 3D Textured Mesh and step 3. DSM, Orthomosaic and Index, if the 3D Textured Mesh and/or step 3 are processed after the surfaces have been created and if the Use for DSM and Triangle Mesh box is selected in the right sidebar of the surface.

Great, thank you!