Book (camera-based) Scanner versus Flatbed Scanner

Built a scanner? Started to build a scanner? Record your progress here. Doesn't need to be a whole scanner - triggers and other parts are fine. Commercial scanners are fine too.

Moderator: peterZ

cbmira01
Posts: 4
Joined: 04 Mar 2014, 00:52

Possible Treventus-type scanner?

Post by cbmira01 »

Hello all, a new poster here..

If you were going to make a handyman knock-off of something like the Treventus Scan Robot,
you might try mounting two wide LIDE sensors in the plunging head movement, angled slightly.
Then the rest of the design would be something like building two flatbed scanners or contact image
scanners in parallel, with two separate data paths (one for each facing page), but on a common arm
movement incorporating a vacuum system. I imagine you'd want to shoot for ~600 dpi resolution.

The problem with this scheme is that commonly-available LIDE sensor heads are 216 mm wide!
This might be fine for some applications, but I'll bet something closer to A4-long (about 300 mm
wide) would satisfy most applications. I'd hate to contemplate the alternatives:
  • * gang sensors together along the width, and deal with inter-sensor seams;
    * move the sensors axially to make up the width, and stitch the data;
    * scan the book more than once and stitch the data;
    * a possible optical solution to cram a 300 mm scan width onto a 216 mm sensor;
    * or, just live with the width limitation.
I'd like to hear your thoughts on a handyman Scan Robot.. Take care!

-- Cal in Louisville KY
User avatar
jck57
Posts: 376
Joined: 23 Nov 2009, 15:21

Re: Possible Treventus-type scanner?

Post by jck57 »

cbmira01 wrote:Hello all, a new poster here..

If you were going to make a handyman knock-off of something like the Treventus Scan Robot,
you might try mounting two wide LIDE sensors in the plunging head movement, angled slightly.
Then the rest of the design would be something like building two flatbed scanners or contact image
scanners in parallel, with two separate data paths (one for each facing page), but on a common arm
movement incorporating a vacuum system. I imagine you'd want to shoot for ~600 dpi resolution.

The problem with this scheme is that commonly-available LIDE sensor heads are 216 mm wide!
This might be fine for some applications, but I'll bet something closer to A4-long (about 300 mm
wide) would satisfy most applications. I'd hate to contemplate the alternatives:
  • * gang sensors together along the width, and deal with inter-sensor seams;
    * move the sensors axially to make up the width, and stitch the data;
    * scan the book more than once and stitch the data;
    * a possible optical solution to cram a 300 mm scan width onto a 216 mm sensor;
    * or, just live with the width limitation.
I'd like to hear your thoughts on a handyman Scan Robot.. Take care!

-- Cal in Louisville KY
An idea for utilizing a page turning system similar to the Treventus adapted to the Standard Scanner:

The platen is hinged so that when it is lifted from the book the two sides of the platen fold up to 60 degrees. Now the platen more closely resembles the Treventus scan head. The vacuum and compressed air features of the Treventus shouldn't be that hard to replicate in order to turn pages.

The US patent application for the Treventus can be viewed here:

http://www.google.com/patents/about?id=lDGvAAAAEBAJ
User avatar
daniel_reetz
Posts: 2812
Joined: 03 Jun 2009, 13:56
E-book readers owned: Used to have a PRS-500
Number of books owned: 600
Country: United States
Contact:

Re: Book (camera-based) Scanner versus Flatbed Scanner

Post by daniel_reetz »

Welcome to the forums, Cal! One of the bigger issues is just the hardware you're starting from. Hacking the GRIN arrays on the LIDE scanners is not easy. I've tried and it's painful,though doable if you are skilled enough with microcontrollers. I didn't have the patience to defeat all the self-checks in the system.

I'd encourage anybody to try this, but won't go down this path myself due to the extraordinary effort to get even baseline access to the imaging array. The moment you'd have the LIDE scanner hacked, they'd discontinue it, leaving you with a brick/something non-reproduce-able..
User avatar
Antoha-spb
Posts: 89
Joined: 21 Nov 2009, 09:54
Number of books owned: 1000
Location: Saint Petersburg
Contact:

Re: Possible Treventus-type scanner?

Post by Antoha-spb »

cbmira01 wrote: try mounting two wide LIDE sensors in the plunging head movement, angled slightly.
Then the rest of the design would be something like building two flatbed scanners or contact image
scanners in parallel, with two separate data paths (one for each facing page), but on a common arm
movement
Then the question is how to synchronize them as each has got it's own step motor driven by it's own chip.

One solution is to use handheld scanners like this one but then the question is about how correctly does the scanner register the movement.

Seams between scanned stripes aren't a problem, if geometry of the scanned image is perfect. All wide-format scanners have these seams due to number of CCD or CIS arrays used, but they have single feeding mechanism and process the CIS/CCD data on the fly by with their firmware, so the user gets a perfectly glued picture.

Another thing about books is the tough binding where you cannot approach close enough to the inner margin with a standard CIS taken from a flatbed/handheld. If you use CCD instead, it will have bigger DoF and can be placed a bit remotely, but anyway you'll need to focus that, what means to hack the scanner completely...

Don't remember if i posted it already on the forum... one person from PA has built a scanner from the available components and made necessary drivers on C++ himself
User avatar
daniel_reetz
Posts: 2812
Joined: 03 Jun 2009, 13:56
E-book readers owned: Used to have a PRS-500
Number of books owned: 600
Country: United States
Contact:

Re: Book (camera-based) Scanner versus Flatbed Scanner

Post by daniel_reetz »

Oh, that's Tom Sharpless! He's a really great guy. When Matti and I were working on a scanner-cam project, he gave us lots of great advice. I really admire him.
peyre
Posts: 3
Joined: 04 Mar 2014, 00:53

Re: Book (camera-based) Scanner versus Flatbed Scanner

Post by peyre »

Hi everyone. I'm looking into buying or building some kind of book scanner myself. My needs are pretty basic--I just want to be able to scan the occasional book without damaging it, and still get good-quality scans. I was looking at flatbed book scanners recently (the ones that take the scan head nearly to the spine), but they're awfully pricy for my needs. Does anyone have experience with those, btw? Do they actually get close enough to the spine to give good scans of most books?

But barring some big influx of money that would encourage me to spend $250+ on something I don't really need, I'm starting to think about building a simple DIY setup that uses a digital camera to scan pages with the book opened to 90°. I have an old camera I could dedicate to this purpose. It seems a fairly simple thing to do, as long as I can work out the details. Am I missing something?
User avatar
daniel_reetz
Posts: 2812
Joined: 03 Jun 2009, 13:56
E-book readers owned: Used to have a PRS-500
Number of books owned: 600
Country: United States
Contact:

Re: Book (camera-based) Scanner versus Flatbed Scanner

Post by daniel_reetz »

Honestly, that's pretty much it. You need a way to support your camera, and you need a light. I mean, check out this recent work by ChrisG -- doesn't get any simpler or cheaper than that, and the final results are decent.

I think I have personally shown that you can make the technology as complicated as you want -- but for occasional, fast scans, you really just need BookScanWizard or ScanTailor, one of the free binders here (DJVUbind, PDFmaker) a camera, and a light.

If you're going to do a whole shelf full of books, the bigger rigs make sense. Don't forget that there are literally dozens of people waiting to buy one at any moment. If you build a good one, you can resell it and get your money back. Then it cost you only time.
peyre
Posts: 3
Joined: 04 Mar 2014, 00:53

Re: Book (camera-based) Scanner versus Flatbed Scanner

Post by peyre »

Thanks Daniel! It's kind of exciting to think I can build a simple rig myself (though the wife may object to me cluttering the office with yet another piece of equipment). Since I have an old camera lying around, I should be able to do this for next to nothing. A couple quick questions though:
  1. I was wondering how to hold the page down for scanning--I was thinking about finding a pair of (somethings) with a 90° bend to them. But am I reading Chris's post right, that he used a small pane of glass? If so, I'd need to change the angle of the light, wouldn't I, to avoid glare?
  2. What resolution should I set the camera at for most books? With the flatbed, I've set it at 150 or 300 dpi depending whether there were pictures or other details that needed to be easily legible.
  3. So in effect, all I need is something to hold the book at 90°, something to hold the camera, and a light source at the right angle?
User avatar
daniel_reetz
Posts: 2812
Joined: 03 Jun 2009, 13:56
E-book readers owned: Used to have a PRS-500
Number of books owned: 600
Country: United States
Contact:

Re: Book (camera-based) Scanner versus Flatbed Scanner

Post by daniel_reetz »

Check out this link from the frontpage -- Matti and I worked on a very simple scanner last year (think of this as a set of recommendations/guidelines and not a step-by-step).

Always capture at the highest possible resolution and/or maximum quality afforded by your camera.

3. Yes.
peyre
Posts: 3
Joined: 04 Mar 2014, 00:53

Re: Book (camera-based) Scanner versus Flatbed Scanner

Post by peyre »

Wow, detailed instructions for building one of these with minimal equipment? That's great! Thanks.

Now comes the hard part, I expect...
Post Reply