TTG's bodged-up, botched-up book scanner

[ThatTallGuy]

I've been planning and brainstorming on this project for a couple of weeks now, and it's time to start doing...

My design is uber-simple and based on a couple of desired goals:
- It has to be doable by me, with my limited (and that's generous) mechanical skills, tools, and space.
- It has to break down into fairly small bits for storage.
- It has to handle books from mass market up to large reference or coffee-table sized.

Most of the pieces roughed out in cardboard or 2x4's in the attached. The top cross piece will be lighter (probably 1x2) and will nest in grooves in the top of the side columns. The side columns will (read: "may someday") hinge to fold in half outwards (meaning the cross-brace will keep the columns from folding during use. The bottom feet are T's whose stems swing in for storage; the columns would fall inward except, again, for the top brace, and the clamps keep things stable during use. There will also be a counterweight on a pulley to nearly balance the weight of raising the platen. The final height of the side columns will be determined by how high the platen goes and how low whatever light I eventually get hangs.

IMGP7633.PEF.jpg (234 KiB) Viewed 807 times

I picked up a couple of broken A590's on Ebay: one doesn't flash, the other has some teeth missing in the zoom gears, but once it's zoomed to the right distance it's fine. Total price ~$125. I'm buying one of Daniel's hardware kits to trigger them.

I found the cardboard-and-lumber mockup to be very useful for good visualizing. A couple of things that the mockup made clear to me:

- Making the bookrest first will let you keep an eye on ergonomics (in terms of locating the platen handle, and how high you have to lift the platen to get under it to turn a page.) The one I ended up with here is several inches taller than my original plan to accommodate my height; I ended up raising the cameras and the lights too to compensate.

- It's going to be helpful to be able to leave the book as near as possible to the front of the bookrest; the platen won't have to be raised as high. This also means that the lack of rigid connection between the columns and the base is a good thing; the cameras can be moved forward to center smaller books when they're at the front.

- A 90 degree angle in the bookrest seems to be the standard around here but it seems very risky to the pages of the book as you lower the platen; the platen could pull on pages as it lowers and crease them inwards near the spine. I don't think a larger angle will hurt the spine -- probably closer to 110 or 120 will be my final opening angle. I expect this will require that the camera be focused pretty closely to avoid picking up the nearer page, but that's a once-per-book adjustment so I don't expect it will be too bad (especially with the mobile columns.)


Comments? Suggestions? Critiques?

[spamsickle]

Looks good.

I'd say your concern about 90 degree cradles is probably unwarranted. I've scanned dozens of books by now, from fat 4-inch hardbacks to stiff paperbacks that wanted to close if I raised the platen too high, and I've never seen a problem with the platen creasing pages. Maybe it could happen if you were scanning a book with onionskin pages, like an old King James Bible or something, but I doubt it even then. And 90 degrees is good for avoiding undesirable reflections from lights that are directly overhead -- direct reflections from one light go from side A to side B and up into the other light, rather than appearing to your cameras. While you could probably place the lights to avoid reflections with a different angle, I'd say unless you have a better reason (really rare books, and a clumsy robot doing the scanning, for example) it would be better to go with the 90-degree cradle and platen for your first iteration.

[StevePoling]

Maybe it could happen if you were scanning a book with onionskin pages, like an old King James Bible or something...

It doesn't have to be an old KJV, I've got a New-KJV (yes, the version is called that) and an NIV that use that uber-flimsy paper, too. (There's a whole alphabet-soup of Bible versions that only differ by some decisions about translating a word here or there.) Happily, if you want the Bible in electronic form, there are lots of free versions out there.

I'm not sure of any other books in common use that use that same kind of paper.

You might say that for book scanning, the Good Book is a Bad Book. Just for grins I tested one of my Bibles with my zeroth-generation platenless system. It was an abject failure. The only book that came close to a Bible in terms of not-working-ness was a 4" thick copy of "The Life and Times of Jesus the Messiah" by Alfred Edersheim(http://www.ccel.org/ccel/edersheim/lifetimes.toc.html). Happily, many hard-to-scan religious books happen to have electronic copies at http://www.ccel.org/

[ThatTallGuy]

And 90 degrees is good for avoiding undesirable reflections from lights that are directly overhead -- direct reflections from one light go from side A to side B and up into the other light, rather than appearing to your cameras. While you could probably place the lights to avoid reflections with a different angle, I'd say unless you have a better reason (really rare books, and a clumsy robot doing the scanning, for example) it would be better to go with the 90-degree cradle and platen for your first iteration.

Hm. Hadn't thought about glare reflections into the camera. The other reason I was thinking about a wider angle was to reduce the reflection from the opposite page as discussed in this thread which mentions significant extra lightening near the spine from light reflected from the opposite page; I thought widening the angle would reduce that problem.

I will try to find a way to build both and see which I like better.

[daniel_reetz]

Yes! Another scanner build! I love the extreme scale, and Cardboard Aided Design.

I'd be very curious to see the results of your wide-platen vs 90 degree platen experiments.

As to the switch we talked about, I haven't found a good DC adapter yet -- Saver's refreshes their electronics stock weekly instead of daily (I'd previously purchased almost 20 5V cell chargers from them -- only 7 were suitable ((where suitable means they output 5V without load))).

[ThatTallGuy]

I love the extreme scale, and Cardboard Aided Design.

Well, I have at least one of the qualities of a great programmer: I'm lazy as all get out. I don't want to have to bend over to use the thing. And at over 6'7", that means either mockups in advance or doing it over. Actually I expect the thing to get smaller; the side columns don't need to be nearly so high, but only building from the bottom up will tell me their correct sizes.

I'd be very curious to see the results of your wide-platen vs 90 degree platen experiments.

I will be sure to post them. Don't expect them anytime soon though. Impatience, I don't have.

As to the switch we talked about, I haven't found a good DC adapter yet

As I said, no rush. In the (very unlikely) event that I finish construction before it gets here, I expect to spend a lot of time in testing and adjustment.

[ThatTallGuy]

An update. Here's what it looks like so far:

_IGP7690.JPG (294.8 KiB) Viewed 699 times

Yes, that's a cardboard mailing tube just resting on top for the lights. It does look like the columns can be shortened, but that will come after the 120-degree platen experiments (the cameras have to be much higher to get a line perpendicular to the face of the book if the book is so much farther open.)

The camera holders were the only thing here which were non-trivial to visualize, for those who (like me) have poor 3D skillz. They have a vertical "tab" to be clamped to a column and a sloped platform for the camera to attach to.

_IGP7687.JPG (46.42 KiB) Viewed 699 times

Note the slot cut 2/3rds of the way down on the right side of the sloping piece. This is to secure the camera. A 1/4x20 bolt (same size as a tripod bolt, screwed into the tripod hole on the camera bottom) and a couple of washers let the thing slot into place. (The washers are spacers to keep the bolt from screwing its way up into the camera innards.)

CameraAttached.jpg (22.63 KiB) Viewed 699 times

Platen and bookrest not done yet; still using the cardboard bookrest mockup. But a test shot with acrylic in place resulted in this:

IMG_0017.JPG (310.41 KiB) Viewed 699 times

Moral: it doesn't have to be fancy to work. If I, with my next-to-nothing carpentry skills, can make one of these things, anybody can.

[StevePoling]

This brings back fond memories of my favorite show of some years back, Junkyard Wars. The ability to use improvised materials in creative ways is one of the talents I admire most.

[ThatTallGuy]

I was always impressed with the Junkyard Wars people too -- but it always seemed a little too convenient that they always found exactly what they were looking for.

So I'm almost there, I think:

Overview.png (401.39 KiB) Viewed 639 times

Off to the side is the monitor, which is my secondary computer monitor which can take composite video input. The black box in the bottom left is an A/B(/C/D) video switch from long ago.

My girlfriend looked at my cardboard bookholder and said, "If it's working, why bother making another one?" I opened my mouth... closed it... opened it again... closed it again.... so now the cardboard prototype is there until it breaks. It does mean that one camera will be higher than the other because the bookholder is not quite cut right, but it will do for now.

Some details:

The platen is just two pieces of acrylic and two end pieces to hold the acrylic in place. The two sheets of acrylic are not attached to each other, just bolted to the end pieces, which are cut from a $3 particle board shelf. I cut 90-degree ones and 120-degree ones but so far I don't have the 120-degree bookholder (it will be simple to make out of cardboard, once I'm a little farther into testing.)

PlatenEnds.png (398.11 KiB) Viewed 639 times

PlatenLifted.png (466.1 KiB) Viewed 639 times

The platen needs to be flexible in height, but only book-to-book, not continuously. So the platen just gets clamped to a hinged column in the back which itself is just clamped to the table. When starting a new book, the platen height (and each of the cameras) gets reclamped as necessary.

PlatenSupportCloseup.png (207.02 KiB) Viewed 639 times

And for moving the bookholder side to side, I dug out some casters which were left over from an earlier project. The casters rotate, however, so to restrict movement to one direction I wedged two of the casters into line.

UnderBookholder.png (175.34 KiB) Viewed 639 times

So I can take pictures now! Well, almost. One of my cameras doesn't want to focus; I think the zoom (which is broken) uses the same gears as the focus, so it just doesn't go anywhere even when manually focusing. Will have to replace it. But other than that... time to start looking at the software.

[daniel_reetz]

As promised, pictures/measurements of the handle assembly (well, not as promised, like two ridiculous weeks late). Things are slowly coming together; my project load + thesis work has me stalled out at a glacial pace. Hope these help. Apologies for the delay.

20100228_008.jpg (356.49 KiB) Viewed 636 times

20100228_009.jpg (277.95 KiB) Viewed 636 times

20100228_010.jpg (333.33 KiB) Viewed 636 times

20100228_011.jpg (344.15 KiB) Viewed 636 times

20100228_012.jpg (286.98 KiB) Viewed 636 times

20100228_013.jpg (288.33 KiB) Viewed 636 times

20100228_014.jpg (327.85 KiB) Viewed 636 times