The Fool wrote:
>
> http://marshallbrain.com/robotic-nation.htm
>
> Robotic Nation
>
> by Marshall Brain
<great snippage>
> The Vision Thing
> One of the key capabilities limiting robotic expansion at the moment is
> image processing -- the ability of robots to look at a scene like a human
> does and detect all the objects in the scene. Without general, flexible
> vision algorthms, it is hard for a robot to do much. For example, it is
> hard for a blind robot to clean a bathroom or drive a car. Part of the
> problem is raw CPU power, but that problem will be solved over the next
> 20 ro 30 years because of Moore's law. The other part is a software
> problem. We don't have really good algorithms yet. My prediction is that
> we will see significant progress in the image processing field over the
> next 20 years.
How much progress has been made on the visual processing problem in the
last 20 years? What is the current rate of progress? How many people
are working on this?
This is the biggest stumbling block to the problem.
Also, how much longer will Moore's Law hold? It gets to where part of
what's making things faster is that the size of components on a chip are
shrinking; there's some finite limit to that beyond which shrinking is
impossible. Then we have to use other methods on the same hardware to
increase speed, or go in a totally different direction with it.
I'm just wondering where the technology is now and at what rate it's
improving.
Julia
_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l
