Hi guys,
I was testing the example projects from pix4dmapper, especifically the Quarry one, but I got high erros in my absolute geolocation. Why is my absolute RMS error so high even though I added GCPs?
Did I mark them wrong? In order to check for accuracy I should check the RMS error in the absolute geolocation section, correct?
I attached the final report down below:
Pedreira_report.pdf (2.0 MB)
I appreciate any help. Have a great day!
As far as I can see there is nothing wrong with the result.
The biggest Error X/Y[m] for the Ground Control Points is less than 23mm. Just fine.
Error Z[m] of 129mm is not outrageous either.
The Geolocation Bias -0.331006(X) 0.588261(Y) 0.072632(Z) is just a shift, a translation between initial and computed image position in the output coordinate system.
Thank you for your reply! Do you know how reliable is the accuracy in the quality report when using NTRIP or GCPs? Are they approximations?
Thank you in advance!
Hi Renato,
Mapper will calculate the errors to the GCPs. These are not approximations but very reliable numbers. As for the NTRIP, these will correct relative to the closes basestation. Typically a rover should be able to acquire a point that is 2-3cm. However, there are many factors that can impact the accuracy such as the distance to the nearest basestation (amongst others).
Hi Mike, thanks for replying to my post.
I did a test with a mini 2 and the results for absolute Geolocation RMS error in the final report is x=86,7 cm, y=89,6 cm and z=63,91 cm. I have attached the final report below:
Galpao_mini_rapid_report.pdf (2.1 MB)
Even though the GNSS receiver on the drone may generate meters of error. Could I say those measurements have been ajusted and are reliable without GCPs on the final report above?
Second question, If I add GCPs or an RTK module with Ntrip and they show a RMS error of just a few centimeters (3-10cm) after processing, could I tell my client that this is the error if he makes any measurements in my model?
Thank you for time helping me understand the final report.
Have a great day!
If the drone is not RTK and you do not have GCPs, then you cannot assess the accuracy.
If you have GCPs and your RMS error is 3-10cm then you can tell your client that the model is reliable within these tolerances.