XMP infos details

Development Asset Creation
,
1

Hello!

I have a technical question about the xmp files.
Can someone explain in detail what each line means? It will really help us with our calibration.
Thank you very much :slight_smile:

<x:xmpmeta xmlns:x=“adobe:ns:meta/”>
<rdf:RDF xmlns:rdf=“http://www.w3.org/1999/02/22-rdf-syntax-ns#”>
<rdf:Description xcr:Version=“2” xcr:PosePrior=“locked” xcr:ComponentId=“{94BFD0F2-3584-4C31-BA1A-B509916FAC8B}”
xcr:DistortionModel=“brown3” xcr:FocalLength35mm=“27.563258135136”
xcr:Skew=“0” xcr:AspectRatio=“1” xcr:PrincipalPointU=“-0.00859425828470126”
xcr:PrincipalPointV=“0.00147661545482012” xcr:CalibrationPrior=“locked”
xcr:CalibrationGroup=“-1” xcr:DistortionGroup=“-1” xcr:InTexturing=“1”
xcr:InColoring=“0” xcr:InMeshing=“1” xmlns:xcr=“http://www.capturingreality.com/ns/xcr/1.1#”>
<xcr:Rotation>0.645631530233224 0.727223280655768 0.233036965395983 0.749656338371684 -0.661719469706966 -0.011946453444071 0.145517358098152 0.182410645201957 -0.972394495567072</xcr:Rotation>
<xcr:Position>-1.73269606826341 -2.61812602738815 31.1784418885721</xcr:Position>
<xcr:DistortionCoeficients>0.193812211964393 -0.439666902222318 0.192149639602353 0 0 0</xcr:DistortionCoeficients>
</rdf:Description>
</rdf:RDF>
</x:xmpmeta>

2

Nobody knows? :slight_smile:
I can’t find the answer inside the help of CR.
It would be really helpfull

3

Hi Karim Toubal
Can you please be more specific what you need to solve ? as there could be a more efficient way to solve your case…

4

Dear,

We have two issues to solve:

The first one is that we have a fixe-rig with 260 cameras and would like to enter an initial position & orientation for each one of them but would like this position & orientation parameters to be used not as the definite position & orientation but to be used as a starting point for calculation of the real position & orientation and that each time we run a 3D-reconstruction. This is to help the reconstruction algorithme to converge faster on one hand and to help the finding of matching points from one picture to the other on the other hand. We have tried to input directly all the xmp files from a calibration set of pictures shouting a calibration target but even if it works fine at first, after some-time with temperature and/or hydrometric changes, some little tiny changes in the position & orientation generates 3D reconstruction artefacts or bad quality reconsrtuctions.

The second issue we are trying to solve in the same time is that today the matching point phase (step one of the reconstruction process) is done using the name of the pictures I assume, hence every picture is scanned for matching-points against only 2 pictures : the one before and the one after; but in our fixe-rig each picture could be scanned for matching points against at least 8 pictures (Right, Left, Top, Bottom and Diagonals) and may be more. What we would like to do is either:

  • Given the position & orientation of every cameras; tell to the software to select the adjacent cameras

  • Or in a simpler way give as input in each xmp files the name of the picture to scan the current picture against.

  • Or even better, input in a meta-file the list of the cameras that needs to be scanned against one another and the number of components for this reconstruction

Looking forward reading about you. Take care :slight_smile:

5

Hi Karim Toubal
Contact me at milos.lukac@capturingreality.com for a discussion on how to use the XMP workflow the right way. So, in short, there are already good and working workflows without the need to care too much about anything…