Elevation Offset Issue with GCPs

PIX4Dmapper PIX4Dmapper Questions/Troubleshooting
1

Hi everyone,

I’m currently working on a multispectral project and I’m facing a persistent elevation offset between my GCPs and the drone imagery in Pix4Dmapper.

Here is the setup I used:

  • Drone: DJI Matrice 300 RTK
  • Camera: MicaSense RedEdge-P Dual
  • GCP collection: Spectra SP85 GNSS receiver (base and rover setup)
  • Coordinate system used in the field:
    • Horizontal datum: NAD83(CSRS)
    • Coordinate system: MTM Zone 8
    • Geoid model: HT2.0 (HT2_1997_GEO.GGF)

I collected GCPs using a base-rover method. The base was placed on a temporary point (not a known geodetic control point), and left running for a few hours to improve accuracy through averaging.

In Pix4Dmapper, I selected the coordinate system NAD83(CSRS) / MTM zone 8 for both the images and the GCPs. However, I’m seeing a consistent vertical offset in the model, the elevations from the GCPs don’t match the point cloud or the generated surface properly.

image
image570×637 81.6 KB

(Don’t mind the quality of the point cloud, I’m still trying to find the perfect flight parameters for our study area)

What we’ve already tried:

  • Verifying the geoid model: HT2.0 is what we used during field collection, but Pix4Dmapper doesn’t allow direct selection of a geoid model.
  • Manually converting orthometric heights to ellipsoidal heights using the formula:
    Ellipsoidal height = Orthometric height + geoid separation
    Based on our geoid separation (~ -31.683 m), but applying this correction didn’t resolve the issue.
  • Reimporting the GCPs with adjusted elevations, but the offset remains.

It seems like a vertical datum mismatch, but I’m not sure how to properly align everything within Pix4Dmapper. Has anyone encountered a similar problem using GCPs collected with a GNSS base/rover and images from the RedEdge-P?

Any help or suggestions would be really appreciated!

Best regards,
Lea

2

Hello there,
It looks like the issue is related to the defining geoid offset for the image and GCPs. PIX4Dmapper doesn’t support the geoid that you are using so you need to calculate the offset and use it. For more information, How to define Pix4D outputs with respect to a Geoid model

As you are using GCPs, you don’t need to worry about the input image coordinate system or the offset value because your project will be aligned based on the GCPs. However, you will see the offset between the GCPs and the images after processing step 1 and it shouldn’t be there after adding the GCPs and marking them. If your GCPs is defined properly, your output result won’t have issue of any offset.

In some cases, it might be difficult to mark GCPs because no images might be available to mark. In that case, follow the steps in the below support article, How to add / import GCPs using the Basic GCP/MTP Editor - PIX4Dmapper

However, if you are correctly defining the coordinate system and using the correct offset value for the raw images and the GCPs, it should align.

2 Likes
3

Thank you for the link, it worked!