US Survey Feet in the Arbitrary Coordinate System Option!

Rosanna,

Feel free to reach out to me for a Teams meeting. I can share my screen to show you where the issue is and explain why it’s a problem.

Pierangelo,

I deselected the solution checkmark because I don’t think that the problem has been resolved. I truly believe that this issue should be resolved within Pix4d Mapper.

To this point, it has felt; for a while now, that Pix4d Mapper is not receiving new tools and updates… Almost as if it is being phased out. There are multiple features that have been requested over the years and have been seemingly ignored, or at least deprioritized in regard to the development of web-based services that Pix has introduced. I’ve been using Mapper since the Pix4UAV days and am saddened by the loss of momentum.

If my analysis is incorrect, please feel free to contact me.

1 Like

Good afternoon @Chris_Putnam,
for my understanding of the replies posted by @Rosana_Lin , she is actually confirming that USft are Supported Units both for GCPs and Images in Pix4D Mapper: honestly it is not so intuitive (first you need to select a generic “ft” in Units, and only when you start to search for a Known Coordinate System, you will finally find some of them in USft), but it is possible.

Attached a screenshoot of the above mentionned option:

However, if what your are experiencing is a lack of Automatic Recognition of the Coordinate System in USft just when you are importing the images, in that case I’m not able to confirm that, because I live too far from a USft Coordinates Zone.

Sincerely,
Giancarlo

What about geoid support? This has not been addressed. And please don’t tell me to use Pix4DMatic. I have a perpetual license for Pix4DMapper. Either keep developing Mapper or deprecate it. You can’t have it both ways. Why am I paying for maintenance if the software is not being developed? It’s frustrating.

1 Like

@giancarlo.patierno, I’m specifically referring to the arbitrary coordinate system option that I have to use for local coordinate systems. That is where the lack of USFT as an option is problematic. I’m not referring to known systems.

I concur. I pay $750/yr for support and maintenance. The low bandwidth in my region necessitates on-prem processing. I don’t want to be told to use Matic to resolve the problem either.

1 Like

Glad to see I’m not the only one sharing in this frustration. I’ve been requesting the ability to use different (non-global) geoids for nearly 7 years now. The only workaround they’ve suggested was to use another program to convert the coordinates. That’s all fine and dandy, I then enter into utilizing arbitrary input and output coordinate systems, which works as long as you’re using an XYZ file for exports instead of an LAS/LAZ, and you don’t mind your orthoimage being by 6’ off when working in state planes. Using this program seems to be a vicious cycle of workarounds, definitely not where I thought it would be at by now when I started using it back in the Postflight Terra3D days.

1 Like

No doubt.

I finally purchased Pix4DMatic, and am a bit dissapointed that a coordinate system can’t be set in Matic without importing GCPs, which is unfortunate because I have clients who just want the imagery or want to look at their roof in 3D and don’t want to pay for survey, but I can’t convert it to a CS using ft or usft without importing control in Matic or Pix4DSurvey.

Hi @Chris_Putnam , you do not have to import GCPs into PIX4Dmatic to change the coordinate reference system. Look at the “pen” icon, this lets you edit the CRS of the GCPs (even without there being any!) which will set the project CRS too (bottom left of the interface):

This is the window you get when hitting on the “edit” (aka little pen) button:

There is an * that mentions that this sets the project CRS too.

Hope this clarified the situation.

2 Likes

Thanks @Pierangelo_Rothenbuhler, I figured that out. However, I have been frustrated with the GCP import. In Mapper, I would find all of the points once the primary point cloud was generated and tag them. I would give them the same name as the GCPs in the csv. When I imported the CSV, it would ask me if I wanted to replace the current information with the information on file. That isn’t the case with Matic, which is unfortunate.

@Chris_Putnam I have an assumption on why you proceed like that, but I will still ask. Why does the standard workflow not work for you?

The steps would be something like:

  • Drag & drop (or select) the GCP file into the tie points table
  • Define the coordinate reference system
  • Define the column names (either manually, or by applying a template)
  • Import and view the GCPs in the tie point table and in the 3D view (as well as 2D view)
  • (optional) Mark BEFORE calibration, but the alignment will likely be less good, except if it’s an RTK project.
  • Run the calibration
  • (optional) Run AutoGCP marking from the menu bar. Note: this works only if you have known targets (cf https://support.pix4d.com/hc/en-us/articles/4402423894545-Automatic-target-detection-AutoGCP-algorithm)
  • Select a GCP in the tie point table or in the 3D view
  • Mark the GCP in the associated images, until all reprojections (the green crosses) are at the right spot in the images.
  • (optional) use the “Auto-mark” helper, if you do, still check that the marks are at the right place.
  • Once you marked all your GCPs. Run Reoptimization (and optionally chain other steps such as the densification, e.g. if it’s the end of the day and you’ll see the project the next day only anyway).
  • Verify the reprojection errors and the errors on the GCPs in the tie points table (order them by values in the errors columns by using the arrows next to the column name). Check that there is no error too large according to your specs.

Okay, that’s about it. Can you please highlight why your workflow is more adapted for you? and what’s the flaw in the one above? That will help us adjust if needed.

Thanks!

@Chris_Putnam just saw this other post: Manual Tie Points
Maybe your workflow is the same? The fact it is in arbitrary coordinates would explain why your workflow works better.

Your assumption is correct. But let’s go through the exercise to make it valuable to anyone who might be searching for an answer to this.

  • Drag & drop (or select) the GCP file into the tie points table

**Pointless (haha), because in an arbitrary coordinate system, they aren’t referenced.

  • Define the coordinate reference system

**I define the arbitrary coordinate system first (which, as an aside would be nice as a template option)

You should add orange Home Depot bucket lids with Gorilla Tape crosses to the known targets list. :grin:

  • Select a GCP in the tie point table or in the 3D view
  • Mark the GCP in the associated images, until all reprojections (the green crosses) are at the right spot in the images.
  • (optional) use the “Auto-mark” helper, if you do, still check that the marks are at the right place.
  • Once you marked all your GCPs. Run Reoptimization (and optionally chain other steps such as the densification, e.g. if it’s the end of the day and you’ll see the project the next day only anyway).
  • Verify the reprojection errors and the errors on the GCPs in the tie points table (order them by values in the errors columns by using the arrows next to the column name). Check that there is no error too large according to your specs.

The rest of the steps are synonomous.

My issue is that, when the arbitrary coordinate system is used, all georeferencing goes away. I usually (in mapper):

  • Define the arbitrary coordinate system
  • Run step 1
    *Import GCP file into Global Mapper and convert it to international ft.
    *Export .csv control file in international ft.
  • Open Google Earth or Global Mapper and import the GCPs there to use as a visual reference to find the targets.
  • Locate and name the targets using the same nomenclature as the .csv control point file.
  • Import the GCPs, at which point Mapper asks me if I would like to replace the current values with those in the file. I replace the current values with those from the file.
  • Re-run step 1
  • Check error
  • Re-match / optimize or continue to steps 2 and 3.
  • Export
  • Open in Global Mapper as international ft
  • Convert to USft
  • Import original control file (in USft) as a final check before distribution
  • Export all deliverables in USft.

What would really be the gamechanger would be a process that allowed me to translate the arbitrary coordinate system to the “world.” In addition to the NEE or XYZ coordinates, there should be columns for Lat, Lon, Elevation where:

X=Lat
Y=Lon
Z=Elevation

That would solve alot. I would still like to have the option to replace the values by importing a new file. There have also been times when I have imported the wrong control file. In mapper, no big deal. I just had to import the correct file and replace the values. That would still be nice.

mhm, ok, I think I get your workflow. You process the project as if it was in a known CRS, there you mark the GCPs. That way you have the normal GCP marking workflow with cameras, GCPs, etc…at the right place. Once marked, you say it’s all in arbitrary, at which point you import the arbitrary GCP coordinates, which overwrite the automatically defined ones from the initial marking. Is that it?

In theory, you could do the same in PIX4Dmatic, except that I don’t think we overwrite existing values of GCPs when you reimport it. Also, changing the CRS in PIX4Dmatic removes any processing, so you would need to start from scratch. So checking whether we can be more flexible with the CRS change, as well as overwriting existing coordinates could be interesting.

Another idea. Would the site localization workflow be another alternative that provides a link from the arbitrary to a known crs? You would measure known points in the arbitrary CRS and in a known CRS, this allows you to create a transformation between the two coordinate systems, which you can import into PIX4Dmatic. These two articles may be interesting for you:

Haha! Let’s see what we can do.^^

Now more than ever. With rtk drones becoming the norm, it would be huge to use our own local USft systems.
Please make USft available for arbitrary and site calibrations!

Chris have you gotten aroud this?

I might be missing something, but you can set ftUS in the Arbitrary CRS definition:

Or were you looking for something else?

I’m referring to Mapper.

Oh right, sorry I got confused as both software were mentioned throughout the thread. Thanks for clarifying.