Standard deviation of image position when using RTK-GNSS.

Can Pix4Dmatic see the standard deviation of the position embedded in the image using RTK-GNSS?
If not, I need to get rid of the bad position standard deviation images beforehand.

Hi @Kaname , I’m not sure what you’re exactly referring to, but here are some ideas.

In version 1.48.0 of PIX4Dmatic, we added the capability to view the input images’ accuracy values in form of an ellipsoid. The goal here is to check which cameras have a good input accuracy (i.e. got good RTK or PPK data) and which ones didn’t. This can give an idea of the data that is used as input for the processing. We’re currently working on exposing all external camera parameters, so that it will be possible to edit them too, for further control on the inputs to the calibration process.

This can be enabled in the content panel in the settings of the input images, see:

Here is an example of how it looks like:

Is this what you were looking for? If not, can you please explain more what you’d like to have? Thanks!

Your answer is what I am looking for. Thank you!
And is it possible to remove unreliable location images from the calibration? In a future update?

@Kaname in version 1.49 we’ve added the “Cameras” tab, it shows all exterior camera parameters (position, orientation, accuracy) and lets you edit them. The multi-selection of cameras in 3D, or in the table is useful for that. You can also sort them by accuracy and then edit the accuracies accordingly. What I would typically do, is to order the images by the Z accuracy, then I would multi-select all of them up to a threshold and if I do not trust the initial accuracy I would set something much bigger, e.g. 2m. That way they will be much less taken into account in the optimization.

1 Like