GCP Ground Control Point Mapping and Picture Tagging File Uploading Workflow Details

I have three questions with respect to GCP’s:

  1. Must EVERY image that contains a GCP in the picture be tagged or associated in the GCP Interface processing? For example, if I have 5 overlapping images that show the same GCP, must I map all 5, or does an increased count just increase the averaging and accuracy?

  2. I noticed a “floater” non-associated crosshairs during the GCP Interface mapping process that I didn’t create. Why is this automatically put on both my image and the map? Is this somehow intended to be an automated part of the process?

  3. Is it possible to apply a GCP file to a project retroactively, or must the file be created via GCP interface and uploaded in conjunction with the pictures before the imagery is processed?

Thanks, -Andrew

1 Like
  1. Nope. Try to get at least 3-5 good/accurate tags per GCP. Eventually you’re just adding error by manually tagging more (you’re not going to be tagging the exact same pixel each time, try as you may), so you don’t need to tag all 50 images that may contain a given GCP, you know?

  2. Are you using POSM GCPi or GCP Editor Pro? I’ve not noticed this with GCP Editor Pro.

  3. We can’t re-process/align the dataset after the fact (like Warp/Georectification in GIS suites), so it must be present during upload for processing.

1 Like

Thanks for the clarification on points 1 and 3, Saijin.

As it turns out, with respect to the “whack-a-mole” point that pops up that I mentioned on question 2, I just hit “delete” at its very occurrence. It seems that it disappears by itself once you have GCP’s picked for all of the pictures. I can only presume it is an automated feature that either isn’t intuitive for me to use or isn’t fully functioning yet.

Thanks again,
-Andrew

1 Like

A post was split to a new topic: Issue with GCP Not Reconstructing