robot capable of distinguishing individual pages

Translate This Thread From English to

Threaded View
I have a question that could be rather far out, but potentially doable
as well.  I am building a robotic book scanner and I plan on using
some kind of arm to pick up an individual page in order to turn it.
However, because of the nature of pages sometimes wanting to stick
together, it might occur frequently that I end up picking up more than
one page with my arm.  I was wondering if there was a way of sensing
if more than one page was lifted up.  I made a rough measurement and
calculation and determined that the average page thickness is .02 mm,
so my sensor would need to be able to distinguish something that
minuscule.  Is this something that should be done optically, or would
it be better to use infrared or ultrasonic methods for measurement?
Even better would be a consistent method of lifting up a page that
would yield an error rate of 1% and then manually rescan the few pages
that were missed.


Re: robot capable of distinguishing individual pages


http://www.canon.com/technology/canon_tech/explanation/compact_interferometer.html
Overkill?  Maybe.

You could just use OCR to check the page numbers...

Re: robot capable of distinguishing individual pages


Some ideas:

1.  Hold an LED against one side of the page.  Hold a
     sensor against the other side.  Measure how much
     light shines through.  Most paper is somewhat
     translucent, but two pages will block much more
     light than one.

2.  Hold a metal plate against each side of a page.
     Measure the capacitance between the two plates.
     If you picked up two pages, the capacitance should
     be about half of normal.


Look at the mechanism in printers.  They need to be
able to pick up one page at a time from the feed tray.
But they don't work perfectly, and the quality of paper
they deal with is probably better than you will get from
a random book.


Re: robot capable of distinguishing individual pages


    There are many page-turning mechanisms, most of which sort of work.
Nobody has one that works a very high percentage of the time and reliably
self-adjusts to the books being scanned.  The Internet Archive book
scanning people have struggled with this.

    The better mechanisms usually involve a vacuum.  See, for example,

    http://www.freepatentsonline.com/20030172795.html

                    John Nagle

Re: robot capable of distinguishing individual pages


Do you know if anyone has tried using a mild adhesive, like on
a Post-It note?  It seems like that would pick up a page fairly
reliably, and if a second page was detected sticking to the
first, it could be pulled loose with another sticky finger coming
from the other side.

The major downside that I can see is that the adhesive would
wear out after a few thousand pages, so you would need to
have some way to automatically renew the adhesive if
manual intervention is to be avoided.


Re: robot capable of distinguishing individual pages


    Yes.  I found that with Google within fifteen seconds.

    Please do some homework before posting questions.  Thanks.

                    John Nagle

Re: robot capable of distinguishing individual pages


The idea of using an adhesive might work.  I read an article which
described using Polydimethylsiloxane, the main ingredient in silly
putty.  I had previously considered using suction to lift a page, but
there might be the possibility of bleedthrough picking up extra pages
with the suction.  Any suggestions about which of these two methods
would be the best?


Re: robot capable of distinguishing individual pages


--snip--

I probably missed it, since _someone_ must have posted a suggestion
to use spot-vacuum-pressure to lift a page... but I'll second it
anyway. <grin!>

Imagine placing a soda straw on the page and sucking on the other
end. Now imagine using a flexible plastic hose, say 1/8" or 1/16"
with an air pump motor on the other end and valved via a solenoid.
I wouldn't think you'd need all that much suction, unless you were
using 100lb paper.

Want to be sure you only have one page and not two or three? Use a
copy of the same not-appearing-in-this-picture mechanism you used to
guide the first hose against the page to be turned, only touch this
second hose to the _back_ of the just-lifted page. If you can lift
the second "hose" away from the page-or-multiple-pages without
breaking the vacuum on the first hose, you just separated an extra
page from the first. Repeat as necessary.

The details (and variants) are left as an exercise for the reader
(translation: I think I'll step off this limb before I _completely_
cut it off <grin!>).


--
    Do not follow where the path may lead. Go instead where there is
    no path and leave a trail.  -- Ralph Waldo Emerson
--
Frank McKenney, McKenney Associates
Richmond, Virginia / (804) 320-4887
Munged E-mail: frank uscore mckenney ayut minds pring dawt cahm (y'all)

Re: robot capable of distinguishing individual pages


    Vacuum based paper handling systems are common and reliable,
but noisy, which is why they're not seen much in office environments.
Large sheet fed printing presses, check sorters, mark sense readers,
and similar machinery usually have vacuum feeders.

    The robotics part of the problem would be to build something that
looks at the book with a camera or two and figures out where and how
the picking and holding devices should be placed.  That would
be an advance over current technology, which usually takes too
much manual adjustment and watching.

    Here's one made from Lego:

    http://www.geocities.jp/takascience/lego/fabs_en.html


                    John Nagle

Re: robot capable of distinguishing individual pages

On Wed, 01 Aug 2007 23:00:33 -0000, dmehling@keenebroadband.com


The usual suggestion is to use a light vacuum device to lift the
page. To get the other pages to let go of the first page, put an
electrostatic charge on the page being picked up.

Site Timeline