Basic Question re Machine Vision
- Can I ask a really basic, ignorant question re machine vision? (I've been struggling with various books on it, and the question/ issue isn't really, clearly put up front).It's this: computers can't really handle shapes, can they? They 'see' digital images - arrays of points - which are then analysed and compared with highly sophisticated and complicated maths.So they can't really compare shapes? - can't simply literally overlay one shape on top of another to see if they loosely fit in an instant? - can't really, directly handle analog images?Doesn't the brain do that?
- feedbackdroids wrote:
>Then it seems like our philosophies are very much alike. I do assume
> > The interesting thing is though the fact that we need to take a
> > area of activity into account, and somehow that wide area binds
> into a
> > single subjective experience. That alone places pretty strict
> > constraints to our ideas of reality.
> Likewise, with a computer, you can bind together a wide area of
> activity into a single reasoning stream, internal world model, and
> behavioral output, and without worrying about making the thing be
> conscious or have subjective anything. The result is that the robot
> can "do" something, whether or not it can "feel" anything, in your
> subjective sense.
that a robot can be made to categorize reality into sensible things and
it can form a conception about self and it could as a result have a
subjective experience. I just can't explain how, instead I assume that
reality simply is such a place where this sort of turn of events is
So, much like in a computer a wide area of activity "binds" in the sense
that that wide area of activity can be expressing a single "logical
thing", I've been saying that our subjective experience is a result of
the brain expressing reality by its configuration. Even when I imagine
how "a brain is expressing reality by its configuration".
After all, there are some pretty good evolutionary reasons for it to do
so (to be able to draw predictions).