Reference point sticker acquisition problem

[[Mi...]]

This is a weird one. We just got a new ATOS Q 12M unit a couple months ago. Started noticing this problem soon after we got it -- the sensor doesn't pick up reference point stickers very well when they're located in the MIDDLE of the Measuring Volume, but it picks them up great when the stickers are at the BOTTOM of the Measuring Volume. We've had ATOS scanners since 2004 -- we started off with a SO4M back then. We don't recall any issues like this with that unit.  We also have a Scan Box,

See attached for a better description of the problem. It's difficult to create a project when the sensor doesn't want to see the reference points.

Any suggestions?

ATOS Scanner Reference Pt Problem.pptx

[[Sc...]]

We are having a similar problem on our ATOS Capsule system, it seems to randomly not pick up the reference points very well. Brand new clean refence points and the sensor won't pick them up. We have tried everything, adjusting contrast, no spray whatsoever...and still no luck. We usually end up just adding more points until the sensor sees them. Any additional feedback on this topic would be appreciated.

[[Be...]]

Hi,

if you want to calculate 3d points of an observed sticker the program has to find the corresponding ellipses in the left and right camera image of the sensor. How is this done? If we have a barcode around each point we can find them individually in each image and can assign the found ellipses to each other via the barcode information. But if they have no barcode around their center we have to "guess". The found center of an ellipse in the left camera image leads to a possible line in the right image (so called epipolar line) and if we find one unique ellipse near this line in the right image we calculate the corresponding 3d point. But if your stickers are arranged in a grid (like in your power point) we find multiple possible solutions -> therefore we reject the calculation of a 3d point because we can't know if we assign the correct ellipses to each other.

Therefore choose a non-uniform distribution of your points!

Hope this helps?!

And again: If you have still problems regarding a special data constellation please contact your local GOM partner in your country and supply the concrete project for further investigations. The Forum is usually not the right place to handle  problems regarding a concrete a data situation.

Regards,

Bernd