Camera Region configuration question

I’m setting up a tulip workbench with a pick to light function.
I’m currently setting it up at home to test first and im having issues getting the regions to identify a change in set up.

I’m looking to see if I’m missing something to do or if my set up is not optimal.

I’m using Intel RealSense Depth Camera D415.
my table is 1.5ft from the floor.
my camera is 6ft from the table.
I have 3 bin holders: 5cm depth 14.5cm width, 22cm length.

for the first two regions of the screenshot below, when i put my hand in the box, it doesn’t pick up on the change and for the third region is looks like it keeps on identifying a change and I’m not sure why:

I’m using the default change detector settings but even when i change the threshold % around its still not working correctly.
Region 3 is always identifying a change for some reason as you can see…

any tips would be great!

Hi @knowlec , thank you for using Vision. For the first two regions, have you tried changing the upper and lower thresholds to be in the range where you’d want to place your hand? You can change the upper threshold to be 500mm to see that it registers any changes between 1-50cm from the bin surface. For the third region, it seems to be picking up on a lot of noise. Can you change the adaption speed setting to be fast so that it can react faster to noise? Let us know how that goes for you :grinning:

Hi, @knowlec maybe this drawing will be of any help.

@knowlec , Adding to the comments. Thtere is a detailed explanation of the Change Detector parameters in this knowledge base article
If you still have issues, I’ll be glad to get on a call and see how we can help you solve the issue.

Reshef Gadot

Hi all,

Thank you very much for the links and troubleshooting.
I have moved the setup to our tulip workbench onsite to continue our pick to light configuration.

I will review the information above when I’m back onsite next week and I will come back to this thread with an update.

Thanks again.