Forum

Author Topic: Import coordinates for laserscans  (Read 3186 times)

Dieter

  • Full Member
  • ***
  • Posts: 213
    • View Profile
Import coordinates for laserscans
« on: April 18, 2023, 11:53:26 AM »
Is there a way to import the coordinates for laser scans?
You can calculate everything well now, but in my opinion this possibility is missing.

Dieter

Dieter

  • Full Member
  • ***
  • Posts: 213
    • View Profile
Re: Import coordinates for laserscans
« Reply #1 on: April 19, 2023, 06:59:13 AM »
Ok I'll ask differently.

I calculate my scans in another program.
Then I output the coordinates of the scans to the e57 file. I read this E57 file into Metashape as a laser scan.
These laser scans serve as a reference for other images that I want to align with the laser scans.
This all works very well, unfortunately my coordinates from the e57 file are not transferred to the corresponding fields in Metashape. I have to keep working with control points in Metashape, even though the coordinates for the laser scans are already there.

How can I bypass this and work directly with the coordinates?

Dieter

Alexey Pasumansky

  • Agisoft Technical Support
  • Hero Member
  • *****
  • Posts: 14927
    • View Profile
Re: Import coordinates for laserscans
« Reply #2 on: April 21, 2023, 01:50:00 PM »
Hello Dieter,

Do you mean that you have geographic coordinates for the laser scans (opposed to some local coordinates, as usual for terrestrial scans) embed to the E57 file that you are importing?

If speaking about the local coordinates, they are imported when laser scans are loaded (point cloud related to the laser scans have their own transformation matrix), although not shown in the Reference pane for the embed spherical panoramas.
Best regards,
Alexey Pasumansky,
Agisoft LLC

Dieter

  • Full Member
  • ***
  • Posts: 213
    • View Profile
Re: Import coordinates for laserscans
« Reply #3 on: April 24, 2023, 08:45:27 AM »
I think it's the georeferenced coordinates.

In version 1.8.0, the process in Metashape also worked with the help of a script. But in the latest version I can't find a way to assign the calculated coordinates from a third-party program to the scans. When importing the E57, the coordinates are transferred to the "estimated coordinates" fields, but they must be transferred to the "source coordinates" fields. Importing the scan positions with the name of the scan file and the coordinates does not work either, since the scan position cannot be found.

Dieter

Tas

  • Newbie
  • *
  • Posts: 10
    • View Profile
Re: Import coordinates for laserscans
« Reply #4 on: November 03, 2023, 05:56:18 PM »
Just wondering if there has been any follow-up on this. I'm knee-deep in research as I'm trying to fully digest previous related posts (https://www.agisoft.com/forum/index.php?topic=12997.60). There seemed to be a workaround solution established with the quick_layout script (https://www.agisoft.com/forum/index.php?topic=12997.msg58658#msg58658), but it seems like that's no longer working for Dieter.

I'm trying to gauge whether I have any chance of success here - if Dieter is stumped, then I doubt I have any hope!


Edit:  I'm not sure that the quick_layout script is still applicable in Metashape 2.0, but I documented a workaround here:  https://www.agisoft.com/forum/index.php?topic=15274.msg68811#msg68811
« Last Edit: November 04, 2023, 01:02:39 AM by Tas »

Dieter

  • Full Member
  • ***
  • Posts: 213
    • View Profile
Re: Import coordinates for laserscans
« Reply #5 on: January 05, 2024, 10:16:28 AM »
I've now had my first experiences with version 2.1 and I'm trying to combine laser scans with photos according to the following description: "Terrestrial laser scans processing in Metashape 2.1"

I have the same problems as in version 2.0, I can't align the photos with the scans.

No matter whether I first want to align the scans with the photos, or the photos with the scans, it just doesn't work. Identical points are found in the scans and in the photos, enough points are shown in the photos and the scans for the matches, but a final alignment does not take place.

I've tried all the combinations with scan groups, fixed transformations for scans and so on but I just can't get any alignment.

Is there anyone in this forum who has managed to do this?

Dieter

e.spiridonova

  • Agisoft Technical Support
  • Newbie
  • *****
  • Posts: 38
    • View Profile
Re: Import coordinates for laserscans
« Reply #6 on: January 10, 2024, 04:32:19 PM »
Dear Dieter,

Thank you for your test and your feedback.  Please clarify if we understood correctly that many matches were found between the images and the laser scans, but they were not aligned?

If yes, do you have the opportunity to send us your project and the images that you used in the project to analyze the problem on our side? Please send the data to support@agisoft.com.

Best regards,
Elizaveta Spiridonova,
Agisoft LLC

Dieter

  • Full Member
  • ***
  • Posts: 213
    • View Profile
Re: Import coordinates for laserscans
« Reply #7 on: January 21, 2024, 10:10:34 AM »
I've learned a lot in the last few weeks and it's basically working now. As a surveyor, I have a few more questions:

1) Can I align laser scans while setting the position of one scan as fixed?

2) Laser scans are now aligned with robust points. If I now additionally do an alignment with photos and optimize the project, will the bundle block also be transferred to these robust points? So are the laser scans balanced with each other and also with the photos?

3) When aligning laser scans, how can I see the quality or accuracy of the alignment?

Thanks
Dieter

e.spiridonova

  • Agisoft Technical Support
  • Newbie
  • *****
  • Posts: 38
    • View Profile
Re: Import coordinates for laserscans
« Reply #8 on: January 24, 2024, 12:00:02 PM »
Dear Dieter,

1) Unfortunately, there is no such possibility now. But we can offer our script like workaround. You need to open a project with alignment (where you removed lock transform and the external orientation for laser scans was changed). After that, run the script. After running the script, there will be a dialog box where you need to select the source file with the laser scan. As a result, the script will shift the result of your alignment to this laser scan.

2) Yes, the key points of both types are optimized together.

3)We can suggest using the Measure Profile tool for horizontal surfaces (you can find information about this tool in our article -  https://agisoft.freshdesk.com/a/solutions/articles/31000170365) . You can also use markers (specify the markers on two clouds and check the error in the Reference pane). Unfortunately, other methods are not currently implemented in Metashape.
Best regards,
Elizaveta Spiridonova,
Agisoft LLC

Dieter

  • Full Member
  • ***
  • Posts: 213
    • View Profile
Re: Import coordinates for laserscans
« Reply #9 on: January 24, 2024, 03:04:13 PM »
Thank you very much.

Dieter

JMR

  • Hero Member
  • *****
  • Posts: 504
    • View Profile
Re: Import coordinates for laserscans
« Reply #10 on: January 28, 2024, 09:51:05 PM »
Sorry, but the patch does not work at all for me as it should. The reference scan is loaded and nothing changes, the scans continue looking up side down. Am I supposed to do anything to tell the group of scans and the linked images to adopt the transform inherited from the reference scan?

What am I doing wrong?
« Last Edit: January 30, 2024, 09:02:42 PM by JMR »