How to align a base model to Mapper generated model

HI all,

I am using PIX4D Mapper and Survey to create .TIN files of stockpiles from drone imagery. These .TIN files are then compared to a .TIN file of the natural ground surface using Carlson (an AutoCAD addon). This process is done to determine the volume of the stockpiles on a month to month basis.

The issue I am having is that my .TIN from the drone images does not align as it should with the base .TIN. There is an elevation difference of probably 30 ft. It seems to be consistent across the model, as if there is a shift error in all the elevation values.

My first thought is that I am using an incorrect output coordinate system. Would that cause such a drastic elevation difference? I am in Montana, using a NAD 83 system.

Second thought is that my images are not georeferenced properly. We just recently acquired the capabilities to perform our own stockpile modelling, and are currently not using GCP’s. My current data uses only GPS location data. In the near future, I will be using a drone equipped with an RTK system for this project. Will the RTK system be enough on its own to properly georeference the images?

There may be some other probable cause of the elevation shift, I am not sure. If you have any suggestions, please let me know. Thanks

Hi @SPEsurveying

An elevation shift of 30ft is compatible with a wrongly selected geoid model.

Make sure that everywhere the definition is correct.
I can’t guide on how to define your TIN in Carlson, but in PIX4Dmapper, these are the general aspects to consider:

  • Make sure that the coordinate system of the images matches the one that was used to write the EXIF/XMP tags metadata. You might need to verify this with your drone manufacturer or RTK settings (if available)
  • If using Ground Control Points (GCPs), make sure you define the coordinate system accordingly to the system used by the surveyor during the collection.

However, it is still possible that the error is caused by inaccuracies in Z value of the standard GNSS data (no RTK).

In general, RTK system will be enough to have a well georeferenced project, but we still recommend to collect some GCPs anyway and use these as checkpoints to assess the quality of your results.

You can find additional information about relative and absolute accuracy here:

Alright, thanks.
I think it is an inaccuracy in my elevation. I put a GCP in and it appeared above the point cloud.

I am hoping that our RTK drone will help with the inaccuracy enough to not need GCP’s. I suppose if nothing else, we could always use just a single GCP at our launch/landing site to check elevation. Would that be sufficient, or is it likely that the elevation inaccuracy is varied depending on location?

Hi @SPEsurveying,

That is difficult to tell. If you have a valid license, you can open a support ticket so we can look more into details.

Can you eventually share the quality report with us?