Forum

Author Topic: Why the RMSE become negative?  (Read 5909 times)

ken lau

  • Newbie
  • *
  • Posts: 5
    • View Profile
Why the RMSE become negative?
« on: September 26, 2022, 06:49:43 PM »
X error (cm) - root mean square error for X coordinate for all the cameras.
Y error (cm) - root mean square error for Y coordinate for all the cameras.
Z error (cm) - root mean square error for Z coordinate for all the cameras.

this three sections contains negative result in GCPs, is that correct?
it is interesting that the total RMSE remain positive result.
another question: Total (cm): 15 and  Image (pix): 0.6, is it good result for GCPs?  >:( >:(

Paulo

  • Hero Member
  • *****
  • Posts: 1351
    • View Profile
Re: Why the RMSE become negative?
« Reply #1 on: September 26, 2022, 07:15:32 PM »
Hi ken lau,

the total error for each component (X, Y, Z) represent the RMS of all camera X, Y and Z errors. The individual error for a given camera can be negative (X_est  - X_source)....

15 cm on fixed GCPs seems large but if the errors are on the cameras it could be OK depending on what type of onboard GNSS positioning you have...
Best Regards,
Paul Pelletier,
Surveyor