Hello
I am trying to use Photoscan to compute the successive positions of a moving stereoscopic camera rig. I have used the methods suggested in the forum (
http://www.agisoft.com/forum/index.php?topic=9016.0) to automatically add scale bars with the known baseline, hoping to fix in this way the scale factor of the reconstructed trajectory.
Photoscan is able to correctly align the cameras and their relative positions seem right (as well as the reconstructed point cloud). However, I don't get any numerical estimate of their positions. It seems that 3 known camera positions or GCPs are needed in order to define a reference frame and be able to estimate these positions. My problem is that I am exploring an unknown environnement and I don't have any external reference either for the camera position or the observed environment.
Is there any way to obtain the estimated positions in a local reference frame (the reference frame of the first camera for example) without importing any reference position? It seems to me that the knowledge of the baseline should be enough to solve the problem with the right scale in relative coordinates.
Thank you in advance for your help
Alex