Talk:Image segment

"The goal to automatically segment images into semantically meaningful parts is very difficult to achieve." And yet, the smallest bird can do it. Doesn't this suggest that the researchers are possibly barking up the wrong tree? The answer probably is quite simple, once the problem is seen differently. I'm not saying I have any particular insight into this, only that it is observable in nature in creatures with very small brains. Figuring out how they do it might be more fruitful than trying to crack the problem from first principles. Having worked for many years on image processing and also being acquainted with one of the field's prominent researchers, it amazes me still how a seagull can determine from a height of 30 feet or more from the ground what is a nutritious speck and what is merely an inedible fragment of rock, and that's in a moving image! Just thought IO'd mention this, obviously it's not something that belongs in this (or any other) article. GRAHAMUK 06:55, 20 Oct 2003 (UTC)