Compact assembly of platen, light, and cameras: Is this feasible?

[L.Willms]

I imagine to hold the platen, the light source, and the cameras in one compact frame assembly, pictured in a sectional view of this rectangular cuboid, with two square sides, shown in light blue. This rectangular cuboid is turned by 45°. It is build of square aluminium tubes. The two square sides are linked three longitudinal bars (shown in black). Parallel to those longitudinal bars a third bar runs between them on each long upper side, each one holding a camera in its midst, moveable up and down along the upper sides of the square, so that the cameras point to the center of the book page.

The platen made of two sheets of glass or acrylic is show in green, the light source in the top in yellow. I think of a stripe of LED lights. The platen is placed on the outer side of the square only to make it visible; actually the glass (acrylic) should rest on an offset from the lower sides of the square, like this:


Since in all designs shown here the light source and the camera(s) are separate from the platen, I think that behind this design are serious thoughts about why this should be. Or should by design idea be feasible? Please comment!

[duerig]

This is a worthwhile direction to explore. There are a couple of previous designs that are based on cubes that you should look at for inspiration.

http://veesmee.com/fbbs/
http://www.bookliberator.com

The biggest problem you will face is with lighting. With lights so close to the platen, it is likely that you will see lots of glare and reflected images of the bulbs in the scans. So you will want to pick the most diffuse light source you can find. Maybe an LED panel instead of bulbs. And you might still need to move the lights up further away.

Please keep us posted with your progress. I really prefer designs like yours that keep a fixed geometry between camera and platen. It makes for more consistent scans.

-Jonathon Duerig

[L.Willms]

Thanks for your comments!

With lights so close to the platen, it is likely that you will see lots of glare and reflected images of the bulbs in the scans. So you will want to pick the most diffuse light source you can find. Maybe an LED panel instead of bulbs.

I think of one or more LED strips similar to this one

(source: commons.wikimedia.org, the common media store for all language versions of Wikipedia.org.
Klick the image for more information on it)
along the top ridge of this assembly, and a diffusing cloth or other material underneath. Maybe a short reflecting roof.

[cday]

A thought that has been in my mind for some time and might be better discussed in a Lighting thread:

When enhancing photographs of book pages, the limit on what can be achieved in terms of whitening the page, for example, is generally set by the evenness of the illumination; sunlight provides very even illumination because the light is parallel, and ideally parallel light is produced by a [ideally point] light source at the focus of a parabola... Might there be a way with this design to use a flexible white reflective linear surface to reflect the light from your LED strip?

[L.Willms]

Might there be a way with this design to use a flexible white reflective linear surface to reflect the light from your LED strip?

Yes, as I hinted in my previous message. But a parabolic mirror to create parallel light beams would prevent the whole book page to be lighted evenly.

I would rather prefer a diffusor underneath the LED strip.

[L.Willms]

The biggest problem you will face is with lighting.

I also think about how the start and stop of the platen movement causes shocks or vibrations of the cameras, which can cause movement blur, or forces to wait until the camera doesn't move any more.