Forum

Author Topic: Refine solve using camera pose estimations.  (Read 1869 times)

mattias

  • Newbie
  • *
  • Posts: 5
    • View Profile
Refine solve using camera pose estimations.
« on: June 23, 2015, 11:51:37 AM »
Hi!

I have a dataset of images where each image has a relatively good estimate of position and attitude as metadata. (position is off by a couple of percent of the distance between camera and target and pointing is off with a couple of percent of the FOV)

I import the camera data via the bundler input.

when I do so the software treats the incoming cameras as already perfectly solved data. So the small misalignments in position and rotation makes the solve come out really bad.

Is there any way to tell Agisoft that the camera data is a good startingpoint but that it could use a little fine tuning?

(Agisoft throws away too many images when just solving from scratch)

I'm a standard version user but if it is only possible in pro pleas let me know anyway.