1 | initial version |
sarcastic mode on
Haha in this setup, like coffee beans on top of coffee beans? Good luck with that. If you do find a good decent working solution, submit it to one of the next major computer vision conferences like CVPR or ICCV.
sarcastic mode off
No really, I am afraid you are hitting one of the most difficult parts of computer vision and image analysis, that is finding an object with about no texture at all, on top of a stack of the same products. So before giving you any more advices on how to advance, what are the hardware options? Do you have depth sensors? 3D images? or only 2D camera inputs?