Support Website Contact Support Blog

GCP vs Geolocation

Dear Support,

Could you, please, explain a bit in detail the main difference between GCP and Geolocation?

As far as I understand Geolocation (or georeference) is a GPS Tag inside an EXIF data of a JPG image.

GCP is a marking with known coordinates.

How does each affect the processing time?

geolocation are the GPS coordinates of the images (where they were taken), GCPs are coordinates o certain point on the terrain.
if you process without geolocation it will take longer time than with geolocation and also maybe some images are not calibrated or you have problems if you do not use geolocation. GCPs do not have impact in processing time.

Hello Vusal,

It is not needed to have geolocation and neither GCPs, but if you want a good absolute accuracy (the position of the model on the earth), then GCPs are needed or have a very precise geolocation (RTK).

At the same time, the geolocation helps to process faster, in case that the terrain is very uniform, then images will not be misplaced when calibrating or fail to calibrate, and as well, in case that the overlap etc is not optimal, the relative accuracy (position of points related to other points) will be better, so the model will have more accuracy.

So we have some users that do not use geolocation of GCPs, some uses only geolocation, some only GCPs, and some both.

The more close to the best practices, the better since there is always some difficulties on projects, so we suggest, in case that good accuracy is needed, the use of GCPs and geolocation.

Regards,

Hello Miguel,

As you’ve mentioned in your reply “At the same time, the geolocation helps to process faster,…”, can I assume that GCPs marked BEFORE Step 1 will also help to process faster?

Regards,

Hello Vusal,

Image geolocation can reduce processing time because it is easier to match keypoints between neighbor images, as we know the relative position between them.

If you mark the GCPs after step 2, then you need to reoptimize. So, marking the GCPs before step 1 saves you the reoptimization time.
On the other hand, you do not take advantage of the rayCloud for the marking of the GCPs.
Marking the GCPs (or Manual Tie Points) in the rayCloud is easy, fast and accurate as the 3D model is connected to the initial images.

Regards,