So I have found when using WebODM that it does not work well when the images used came from multiple drones but it does work well when all of the images came from the same drone. It seems that the different drones use different elevation systems. Does anyone know why this is the case and if so how one could remedy the situation.
Multiple images of different elevations is fine but the issue is different cameras. When the software processes the imagery it assumes a camera model based on exif data associated with each captured image. If you fly several drones (cameras) the software can’t properly process the image set due to the imagery being from different cameras.
Okay good to know. It sounds like there really is not a work around for this and it would be best to recapture the images using the same drone.
Different cameras definitely work, different elevations work, but back to the original statement/question: if the way the different drones elevation gets recorded is sufficiently different and ODM uses those elevations for mapping, I could see problems with reconstruction. For example, if one drone sets height above takeoff as the Z and the other drone uses absolute (datum or sea level) height, then there could be an issue.
Good to know I assumed the software attempted to create a camera model for all of the imagery and that it would balk at several different camera lens specification because it would try to create a single estimation of the camera parameters for all the imagery.
So provided the exif data from all the imagery has the same X, Y, Z reference it would be able to process imagery from several drones?
Does anyone know how to make it so all of the jpegs of interest end up having exif data with the same X, Y, and Z reference?
Yes. And not just theoretically. I’ve done this many times.
Exiftool is your friend. You can even do mathematical operations while updating Z. Take a look at my examples of setting Z error values with exiftool on this forum. It’s the same process.