Now that the parts gathering for my robot project is moving alone nicely ( with much help from this newsgroup ! ) I am looking into the vision processing I'll need.
I am relatively spoiled in this area since I have been used to using the excellent vision processing capabilities of the Evolution Robotics ER1, which are just incredible. Since I am perhaps the world's cheapest robot hobbyist, I would like to hear from anyone with experience and opinions on cost-effective platforms such as the CMUcam2 (such as from Acroname) vs. going with Logitech's QuickCam SDK and the QuickCams .
Thanks for any input !