Well, the filter kinda sorta worked.. It more functioned as an edge
detect on the areas instead of what I wanted... I'll try a ANN
tomorrow.. My wife Ashley just had a beautiful baby girl (Dakota Rose)
on 5/29 @ 1:57pm so I'm kinda tired out(and so is both of them)

--
andrew

On 5/29/05, andrew kelley <leetslacker@gmail.com> wrote:
> An algorithm would be nice.. But I think I found a way to do it..
> 
> use a filter like this (the laser scans horizantally)
> 
> float filtconsts[9]=
> {-1/6,
> -1/6,
> -1/6,
> .1/6,
> 2/3,
> 1/6,
> -1/6,
> -1/6,
> -1/6};
> 
> multiplied by the vertical pixels.. So if you add all the
> constants(uniform color), you get 0.  if you have a value near the
> middle of the Y, the value falls through (> 0), if its a patch above
> or below center, it comes out <= 0.
> 
> Or I'll train a NN to evaluate it so it will pass \ / - | patterns and
> reject all others..
> 
> andrew
> 
> On 5/29/05, William Chops Westfield <westfw@mac.com> wrote:
> > >>>> What sort of filter would I use to eliminate large areas of uniform
> > >>>> color but keep smaller areas?
> > >>>>
> > Just to clarify, are you looking for an algorithm, or the name of one
> > of the commonly available "filters" (or "effects") in a high-level 
> > program
> > like photoshop?
> > 
> > BillW
> > -- 
> > http://www.piclist.com PIC/SX FAQ & list archive
> > View/change your membership options at
> > http://mailman.mit.edu/mailman/listinfo/piclist
> >
>

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist