Multi-camera setup

[chort]

Hi,
I'm using three synced cameras rigidly mounted on a ROV (remotely operated underwater vehicle) with about 70% overlap at typical survey altitude.
What would be the best way of importing them using a Python script, so they get calibrated as separate cameras, but contribute to the same 3D model?

Also, as the scene needs to be left undisturbed I cannot place scale-bars or measure the objects with divers, but I need consistent scaling of the object. Having a fixed baseline between cameras should help, or am I not getting the operation principle?

Thanks

[chort]

Let me explain the setup a bit further to make potential answers to this question slightly easier.
We collect continuous imagery flow with 5 FPS with all three cameras, and the ROV is flying (mostly) in a straight line, so it is a kind of "aerial survey" scenario. Positioning is also well defined, with excellent, sub centimetre relative positioning and <0.5m absolute positions derived for each photo location, along with survey grade INS providing roll, pitch and yaw of the ROV.

I had good success with loading in the flight logs through GUI, but i would like to make the process a bit more robust for operators that are not so skilled, hence python. I would also like to separate the photos from the three cameras (which are of the same make, model and have fixed focus) into calibration groups for each camera. Is this achievable through GUI, i can find my way around Python interface myself afterwards.

Thanks

[Alexey Pasumansky]

Hello chort,

In GUI you can split camera calibration groups in Tools Menu -> Camera Calibration window. From Python API you can do that by creating additional chunk.sensor instances and re-assigning them to the camera instances.

You can also use scale bars that define the distance between two camera positions (from GUI select two cameras and use Create Scale Bar option). It's better, however, to create those scale bars after the initial alignment is finished.