This sounds like something out of the Flint Stones.  In fact this slate slab counter may well be used in a quarry.  Which implies dust and vibration.  Anything to do with lights may have problems with the dust over everything.  And even the light source/sensor may get cloged with dust.

Also, if my memory serves me right (which it does less and less these days) slate is composed of layers of rock.  And a slab can be split into two slabs easly.  What that means is that the edges will have horzional lines across it where the different layers have ben deposited a few years ago.  And this could look like where one slab meets another slab.

If a measurment of the height of the rock pile would susfice, then it may be possible to hand a rod down in the path of the pile.  As the pile goes under the rod, it pushes rod foward.  As the rod is pivioted above, this just tilts the rod from the vertical.  When the pile gets far enuff along the line that it starts to pass under the rod and the rod starts to drag along the top, the rod will stop changing its angle of dangle for a peroid of time.  This steady angle can be a signal that the rod is at the top of the pile.  Then it is a little bit of math or something to determin the hight of the pile.

One problem may occcure if a real small pile goes through with a real high pile in front and behind it.  And visaversa.

Bill C.  bill@cornutt.com



----------
> I wonder how different is the reflectivity of front and side surfaces?
> Perhaps using polarized light...
>
> I seem to recall that when I've pointed my camera at a house with a slate
> roof a polarizing filter made quite a difference. It might have been wet
> at the time.
>
> To expose some of the front surface of each, they could perhaps be stacked
> in a "hopper" that is not quite square, and leans backwards. As the
> dimension of each slate is not fixed, you'd have to look at the rear
> portion against the back, with a slot cut in the back of the hopper. Aim a
> light source at the "staircase" of slate edges, at an angle, and measure
> the reflected light with a light sensor behind a polarizing filter.
>
> The above is totally untested. Apologies if it duplicates anything
> already said.
>
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=
>
> Dr. Simon J. Melhuish
> N.R.A.L., Jodrell Bank.
> http://www.jb.man.ac.uk/~sjm/
>