4D

anonymous_user_a4d5aafa1 · October 31, 2016, 2:38am

Hi.

Is it possible to reconstruct a sequence of images from multiple cameras? You can do this in Photoscan.

I was wondering if it’s possible.

Cheers

wishgranter042 · October 31, 2016, 7:57am

Hi Clutterbuck
Yes, it is possible. Do you want to use this for full-body scanning or ?

anonymous_user_a4d5aafa1 · October 31, 2016, 6:58pm

Hi.
I want reconstruct a face peformance.

How do I set this up?

Cheers

wishgranter042 · November 2, 2016, 3:05pm

Hi Clutterbuck
So your scanner is properly calibrated etc, so that you can use the XMP workflow ?

anonymous_user_a4d5aafa1 · November 4, 2016, 1:00am

Hi.

If I have multiple frames per camera and multiple cameras are you suggesting I do a separate reconstruction per frame using the same camera XMP calibration? Can Reality capture load a sequence of images for a single camera? Or would I have to script this? Building a project per frame?

Cheers

wishgranter042 · November 4, 2016, 10:50am

Hi Clutterbuck
Yes, use the XMP workflow and reconstruct the model “per frame”. The XMPs will hold the model in the same coordinates, then you just need to export it “per frame” and put together in an external application in order to get it into 4D…

Anonymous_5cadf75156 · May 2, 2018, 10:05pm

Hello - I am wanting to do a similar thing (re: multiple camera videos imported as image sequence, Photoscan’s version of importing multiframe cameras, and building the model for each frame).

Can someone explain to me how this is done in RealityCapture - first, how to import the photos so that you do not have to do each frame manually?

Also for now I only have the demo version, so does this mean I would not be able to save/export the xmp/calibration info? Apologies if this is simple to do, I have used Photoscan for a while and am working on a project using ‘4d’/videogrammetry, and would really like to figure out if it is worth switching to RealityCapture.

Thanks for your help,
Kelly

Anonymous_243cb9bf6e · May 3, 2018, 6:19am

Hello Kelly,

You can use batch processing via command line to automatically reconstruct a model for each frame. XMPs workflow can be used for preserving the coordinate system of all models. However, as mentioned above, you need to stitch those models in an external application.

Demo version supports processing via command line. In the Demo version, you can save your project but you cannot export the results. The comparison of the supported features in all RealityCapture licenses as well as their restrictions can be found here: https://www.capturingreality.com/Product#Compare