- I have camera vision working for 1) face detection and recognition and for 2) landmark recognition which lets the robot pinpoint it's location in a room or on an XY grid map.I'm finding out that navigation and obstacle avoidance requires yet a different set of visual skills...STeven just emailed me this video. I wasn't thinking of using an omnidirectional mirror but I can't argue with something that works! I don't understand the algorithms used but I'll start looking into this....Comments or help from anyone who has an understanding of the math behind this would be greatly appreciated.....Dr. Bruce.----- Original Message -----From: "RoboRealm"Sent: Thursday, November 05, 2009 1:07 AMSubject: Vision Nav> Perhaps this offers a clue as to how to do this ...
> Note that it does get through a door just using vision!!