Scan Match - different results at different work stations

[[Ka...]]

Hello everybody - novice user here.

So, we have a big problem with the matching of 2 objects. We are matching the objects and we got good results. So far, so good. Then we move to a second work station, we are doing everything exact the same (same version of the software) and we get really different results. Really we were a bit in a shock, so i a next step we scan a cube (20x20x20 mm), measured it in real and compare the results from the two diffrent stations. As a result one station give us real results (-0,0975mm) and at the other station the cube is gone (-20.09 mm).

Can anyone please help us to find the problem? Is there a setting problem or is it common to somebody? 

Thanks a lot,

Best regards Karo

[[Be...]]

Hi,

this sounds very strange. I can only guess: Is the matching of your objects not really constrained by shape? For example, if you have nearly a plane area and you are using a Local Bestfit sometimes the mathematical algorithm stops at different "arbitrary" positions (some algorithm are working with so called RANSAC strategies where only a subset of points are sampled as tradeoff to guess initial start values for the optimization problem and therefore the results could differ). In these cases varying the input parameters can lead to completely different results, but the original reason is that the problem is not well defined. Then you should use "other" alignment strategies.

If you have a professional support contract you should contact your local GOM partner and supply the data for a deeper investigation.

Hope this helps?

Regards,

Bernd

[[Ka...]]

Hi,

thanks a lot for the fast reply. We will contact GOM as fast as possible.

 

Regards,

Karo