Forum

Author Topic: Compute Total Marker Error in m  (Read 2003 times)

mschul

  • Newbie
  • *
  • Posts: 2
    • View Profile
Compute Total Marker Error in m
« on: March 30, 2023, 03:02:44 PM »
Dear community,

I am trying to write a script to calculate the total projection error of the markers in m.

I found the following example that computes the individual marker errors in pix: https://www.agisoft.com/forum/index.php?topic=14595.0


However, the conversion from pix to m is not clear to me.  I would be very happy if someone can help me. I am using Metashape 2.0.0.

Thanks in advance !
« Last Edit: March 30, 2023, 04:56:03 PM by mschul »

Alexey Pasumansky

  • Agisoft Technical Support
  • Hero Member
  • *****
  • Posts: 14854
    • View Profile
Re: Compute Total Marker Error in m
« Reply #1 on: April 12, 2023, 04:08:29 PM »
Hello mschul,

Errors in meters and in pixels are not directly related to each other, as they describe different kind of errors.

Error in pixels indicates the difference between marker projections (set by blue/green flags on the images) and the 3D marker location projected back on the same images. Whereas the error in meters is related to the difference between the 3D location based on the Source coordinate values and location estimated by Metashape. Also note that error in meters is calculated differently for control and check points: estimated locations of the control points take into account both reprojection error and measured object space coordinates, while check points estimated locations are calculated based on minimization of the reprojection error only.

So what kind of errors you would like to calculate using the script?
The example of the error computations in meters is available in the following script: https://github.com/agisoft-llc/metashape-scripts/blob/master/src/save_estimated_reference.py
Best regards,
Alexey Pasumansky,
Agisoft LLC

mschul

  • Newbie
  • *
  • Posts: 2
    • View Profile
Re: Compute Total Marker Error in m
« Reply #2 on: June 01, 2023, 03:21:06 PM »
Dear Alexey,

Thank you for the detailed answer and sorry for the late reply.
I was able to solve my problem with the example in the link you provided.


Kind regards,
Max