Georeferencing using GCPs

[dmct]

Hello Agisoft community!

I have a quick and simple question. When importing my photographs and if they are georeferenced and at the same time I am using GCP´s, does photoscan georeference the model using both the photographs and the markers GPS information? Or by using markers does it automatically omit the photographs information?

Thanks in advanced!

[stihl]

By loading geotagged images into Photoscan, the coordinates will be automatically loaded.

After importing and marking your GCPs, and their accuracy far exceeding the accuracy of the images, you will need to disable the camera coordinates. You do this by selecting all the cameras (ctrl+a) > right click > Uncheck.

Then optimize your tie-points. Now only the GCP coordinates will be used in order to geo-reference your project.

[JMR]

If camera coordinates are given by regular in-camera or drone gps then they will be about 3m accurate, on the contrary, GCPs are usually surveyed much more accurately (something about 1-3cm). If the refference settings are correct accordingly, you may just forget about camera coordinates. Their influence will almost be negligible in most cases
You can see on you own what changes if all camera coordinates are unchecked or cleared... normaly nothing  but fractions of mm.
Regards

[stihl]

JMR is correct if you've kept the accuracy at the default values in the reference settings as the camera accuracy is set to 10 m by default and the GCPs (markers) are set to 5 mm. The camera coordinates are then basically neglected.

Adding camera positions will always help with the initial alignment in terms of proceesing speed and overall accuracy.

[dmct]

Thank you very much for clarifying!