Hi,
I am trying to do something similar, but I am having troubles figuring out how to isolate the object (I'm a machine vision newbie). Basically, I've built a color sorting machine with an overhead camera on a conveyor belt that has small stones running through it. They are, for the most part, similar in size, shape, and color, but not identical. So basically, I don't think that I can use geometrical features to identify the objects, and I can't identify by color, as this is the output range I am looking for (Lab channel), but I am probably wrong about that. Essentially I need to identify how many stones there are in the frame (especially if they are touching each other), and if not more than one, would need the Lab channel readout for further processing, otherwise step should fail.
In the attached pic., there are 2 stones in the ROI, one black, one yellow. I would greatly appreciate if someone could suggest which (vision assistant?) steps/configuration I should be using to achieve this?
Thanks!
Jonathan
