Trimming Sparse Point cloud before Dense cloud - does it help?

[Berend]

Hi, in short.
Does the trimming (select and delete unwanted points) in a Sparse Point cloud help, before running the dense point cloud step?

PLUS
Is there any video workflow that you can recommend for: "Form Zephyr ... Exporting mesh... Import mesh into Blender, fixing mesh/stuff in Blender and Importing it beck to 3df Zephyr"

Thanks for great software / worth the buy!!!

[cam3d]

Hi Berend


Beyond setting a bounding box, there isn't much point in trimming the sparse could.

Knu summed it up well on a Steam post:

"...the sparse cloud is the "root" of your project, and thus you can have only one per project. While you can alter it a little bit, consider that the output of the first phase is not just a sparse cloud, but rather the externals of each camera (the orientation so to speak, rotation and position) as well as the internals (focal length, distortion, etc.).

Having little noise in the sparse cloud is not a big deal - what you should check in the first phase is that the cameras have been auto calibrated correctly (this happen most of the time, but may fail when using few and/or images taken with very distorted lenses)- basically, check that the camera positions you see make sense.

The "extract dense from sparse" is very rarely used and for better results you should always process a dense cloud using the MvS algorithm.

So while you can do minor editing in the sparse cloud, it's rarely done - if the orientation is wrong you'll get bad result anyways in the other stages..."


+ Here's a video from Zephyr to Blender:-) https://youtu.be/oMUGxUqAql8

[Berend]

Thank you for the info and your time!!!