Reply to Thread

Message:: [QUOTE="-jeffB, post: 26598868, member: 27832"]I'm inclined to disagree here. We've been able to do image analysis at superhuman levels for many years now, and vision systems that can inspect coins "in-hand" could pick up details at both small and large scales that humans simply can't see. I think image analysis models will quickly become better at identifying die pairs than humans can be. From photos? That's tougher, because image processing inside cameras and image compression for storage throw away details that aren't relevant to human vision. So automated systems won't have quite the opportunity to excel there - but they can still be trained to avoid human failings ("that random pattern of blotches must be a mint mark!"). It'll be interesting to watch, and at the pace we're seeing, a lot more can happen in what's left of my lifetime. (I hope!)[/QUOTE]