Slashdot Mirror
Japanese Researchers Develop World's Fastest Book Scanner
An anonymous reader writes "IEEE Spectrum reports that Tokyo University researchers have developed a superfast book scanner that uses lasers and a high-speed camera to achieve a capture rate of 200 pages per minute. You just quickly flip the book pages in front of the system and it digitizes the pages, building a 3D model of each and reconstructing it as a normal flat page. The prototype is large and bulky, but if this thing could be made smaller, one day we could scan a book or magazine in seconds using a smartphone." The article mentions Google's similar dewarping system; the difference here is speed.
Does it come with a shark-mount?
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
we just need a few at libraries to digitize everything for everyone... no need to make it any smaller.
Oh, my, isn't this going to be a huge copyright scandal in a few years? I could walk into Borders, scan a few books onto my iPhone, and walk right out without paying.
I have a bad feeling about this...
Johnny 5: Alive!
The project uses a high speed camera... so if a camera from a handy is going to be used, they are going to have to get a lot better.
Nothing in the world is more dangerous than sincere ignorance and conscientious stupidity.
Faster method:
Cut the spine of the book off with a bandsaw with a metal cutting blade (finer pitch teeth than typical wood blade)
Run thru sheet feeder scanner twice, once for each side.
A bit of scripting hackery later, one fresh PDF! Or .djvu, or whatever.
For those of us brought up that its sacrilegious to damage a book, realize that many books were printed on acid paper; yellowing, decaying, brittle, and will soon be dust regardless of what you do, so may as well preserve the content and properly recycle the pulp.
The bandsaw trick also works on magazines, you know, the things we used to read before websites.
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
1) Yes, but does it run Linux.... ... the book scans you! ....
2) Imagine a beowulf cluster of these...
3) I can't understand 200 pages/minute, what's that in LOC/furlough?
4) I can't read you insensitive clod.
5) In Soviet Russia, the book scans the book scanner...wait that's not quite right...ah, got it,
6.1) Scan books real fast
6.2) Tie into massive database that indexes every perceivable medium on the planet
6.3) Get sued by publishers.
6.4)
6.5) Profit!!
7) How fast can it build a 3d model of Natalie Portman with hot gritz?
8) The CIA will use this to scan every page of the manuscripts you've stored in your apartment and will come for your tin foil.
9) Netcraft confirms: reading is dying...
10) A book scanner is like a car that drives really fast over a highway full of book pages...
Someone needs to fix the above list for me.
Now if they also will learn to shred the books in the process and sell the technology to Google, then I will really respect Vernor Vinge's insight (Rainbows End)
This guy has produced some really fascinating work, I strongly recommend checking out some more of it if you have some free time. The high-speed robot hand he developed literally made my jaw drop.
Dislike the Electoral College? Lobby your state to join the National Popular Vote Interstate Compact.
Moron. This was Japan.
Some idiots can't even troll properly.
Never going to happen unless you're scanning some tiny matchbook sized books. Half the problem is turning the pages quickly and reliably. Any handheld will be limited by your own mortal finger speed. Besides, once one person scans a book and makes it freely available, it will be easier for you to get their copy instead of scanning your own.
is that it can only scan child-porn manga.
Mr. Sulu would not be pleased...
"Watanabe told me he was particularly interested in scanning manga comics" Why does cool Japanese tech always end up back at tentacle rape?
"I'm not a quack, I'm a mad scientist! There's a difference." - Dr. Cockroach
is soooooo relieved...his warp engines really need a rest from Captain kirk's demands in galactic emergencies...
The article mentions Google's similar dewarping system; the difference here is speed.
There is nothing preventing Google from pushing high speed video through their book software. In fact, they could probably do that with very little work, since you can use an off-the-shelf high speed video recorder and then just push the frames through the regular processing pipeline.
The reason they don't (and nobody else does) is because it's not useful. For getting acceptable quality from book scanning, you need upwards of 10 Mpixels to get anything decent. Even if you had a 10 Mpixel high speed camera, you still need some control over lighting and camera/book angles for decent results.
Two words: shrink wrap.
Mind, considering that browsing at least the first couple of pages (or more, for non-fiction) is often essential to the sale, that may have an adverse impact. (Especially when you consider the added cost - not much per book but it adds up.)
The prototype is large and bulky, but if this thing could be made smaller, one day we could scan a book or magazine in seconds using a smartphone.
You lost me here. How exactly do I scan an entire book or magazine in seconds using only a smartphone. Somehow I imagine this technology is slightly more than software, unless cameras start coming with super-fast automated page turners attached.
There was an episode of Futurama where Bender is captaining the ship, and Fry asks him if he's read the manual. Bender flips through the several-hundred-page book in about a half second and proclaims "Done", then proceeds to quote it.
It always seemed like a plausible thing to me. Isn't that what they're doing here?
I have developed a truly marvelous proof of this comment, which this signature is too narrow to contain.
There was a similar post in december last year. Main difference seems to be speed. That did 400 pages in 20 minutes, this new one does 200 in 1 minute.
Brain surgery - it's not rocket science!
one day we could scan a book or magazine in seconds using a smartphone
I'm an American. I love this country and the freedoms that we used to have.
In Star Trek TNG episode "The Royale" Data reads the paperback novel "Hotel Royale" in seconds by flipping through it.
I believe the narrator in the video says that the high speed camera is scanning 1000x1000 pixels, and the book he is scanning has very large type, with fewer than 20 text lines per page. I imagine that this scanner can't scan normal text as fast as the Google book scanner.
Sure, it can scan 200 pages per minute... but I could swear I saw it's lips moving as it was reading!
I've abandoned my search for truth; now I'm just looking for some useful delusions.
Technology like this will cause the publishing industry to go the way of the music and movie industries.
Right now the publishing industry is where the music industry was 7 years ago. Multiple incompatible book formats, DRM that lets rights holders yank your paid content away from you, DRM/formats that leave you tied to specific vendor readers, etc.
The barrier of scanning a book has made the publishing industry think that they don't need to provide books in a format that users want and feel that they can keep books locked down by DRM.
Even if DRM succeeds in keeping e-books from being redistributed, scanning technology keeps moving forward. All it's going to take is some enterprising company to buy one of these scanners and become the "AllofMP3" of the book world -- selling e-books in open, non-DRM formats for $1/each and the publisher's business model will fail miserably and they'll try doing the same catch-up that the music industry is involved in in trying to give users a reasonably priced legal product that can compete with the cheap illegal copies.
it won't even take new scanning technology for this to happen -- a scanning "peer-to-peer" service can be formed where thousands of subscribers are asked to scan and proofread a single page from new releases, which are them compiled into a central database to form a complete scanned book archive.
Whether they like it or not, the book industry is going to be forced into open, interoperable standards for books, and lower prices that consumers have come to expect from industries where nearly all of the manufacturing and distribution costs have been eliminated by electronic distribution.
Why the fuck are we scanning books? Isn't there, you know, a DIGITAL REPRESENTATION which is used during typesetting? This reminds me of that crazy story of the person who printed out a spreadsheet, scanned it in, printed out the scan, laid it on a wooden table, took a digital picture of it, then uploaded it to his web site (or something like that).
I believe that Google owns the patents on this approach.
I prefer the "u" in honour as it seems to be missing these days.
The fastest non-destructive book scanner.
The fastest are ones where you chop off the binding, run the pages through an industrial scanning machine, and dump the blob off into modern character recognition software.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
To put it in perspective, you'll need over 5000 years to process all 7 million books in the U-Mich library using one of these, or one year with over 5000 such machines, round the clock.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
how long it takes for the authors guild or whatever they're called to brand this as a purely copyright infringement machine.
When established industries become prey for new technology, why do they resist and ask for protection? This is a fundamental question of society. We protect indigenous peoples. We have copyright and patents. We do much to preserve the old along with the new - backwards compatibility. Why do we not simply tell such industries that it's time to change and support them through the change? Yes, I get the whole free market thing, but rather than fight them to force them to accept change, why don't we offer them ideas and methods to change their business model to match the change in consumer requirements?
No, I'm not being trollish or suggesting stupidity. Why can't we crowd-source ideas for how these industries can recover from game changing technology? Must we wait for Jobs to tell us?
It's just a question.
Support NYCountryLawyer RIAA vs People
last remaining copies of works disappear everyday. everything needs to be scanned in. it's pitiful that right now it's basically impossible to find a freely downloadable pdf of the public domain book "the compleat housewife" Eliza Smith 1727 or other such books.
I would not consider that a large and bulky scanner at all. Check out these scanners being used by Robarts Library in Toronto for the Internet Archive project: http://www.flickr.com/photos/15573112@N03/4252089363/in/set-72157623030966641/
Got a CAT scanner sitting around? Stick the book under it and image each page as a thin slice.
I don't really know if this is technically feasible. How does ink react to xrays? How thin can a image slice be? How fast does a CAT scanner run? Maybe someone with CAT scanner knowledge can fill me in.
Oh yeah, we can use a smart phone... all it needs is... LASERS and... a high speed high res camera... I'm sure that 3MP camera with no flash you have on your phone will work out just fine... you just need to duct tape a laser pointer on there.
The good:
The bad:
Get a really powerful computer. Write a program that simulates a billion monkeys typing on a billion typewriters ...
Moore's law says this will soon be the cheaper method
The US government have made it clear that we have no inalienable rights; any we do not defend vigorously will be taken.
your script might be very simple if you use pdftk. It's a very powerful pdf merging and exploding program.
A few months ago I asked my city's transit if they would post pdfs of the schedules on the web page. They print route schedules/maps and provide them in malls, campuses, and larger public places all over the city. Online, they use Navigo trip planner, links to pdfs and gifs of route maps, and text links to the schedules. So obviously they have some graphic designer in a hole somewhere making this stuff, and probably with InDesign.
Despite all the obvious cost in printed materials, and huge effort in the web site, apparently it's too much to ask to export & post a simple pdf of the indesign they used for the printed schedules. My options are to go through every route and print to pdf with their tables that print strangely, or scan the actual schedule (or carry them, but I really need to minimize the junk I carry with me). The other option is to use a data plan on a cell phone. However, this has not stopped them from offering a "hip" TXT service to relate to the youth of today. All I want to do is slap a copy on my desktop and laptop, and one on my Archos for good measure too.
My friend who works at a University had the job of going through archives of "important material" and scanning it in. This "important material" had virtually all come to the email of the computer technical support... as in, they received it digitally, printed it, possibly deleted the email, but now want to scan and possibly OCR it.
I'm breaking my back and backpack from trying to cram CCNA & MCSE books with a binder and a laptop with me. No, seriously, I had minor back problems, but now they are much worse.
I used to work at a print shop where the teacher's society wanted to see if we could scan in the annual reports of the last ten years. They were printed in a green and purple tone with had pretty much the same colour value. They wanted small files (i.e. greyscale). They did not have copies of the indesign/coreldraw/whatever file.
I understand the posted article has some merit with old books, but there are so many times I see at attempt to fix a problem that has solution in the first place. If something is created with a digital file, then use it, for crying out loud. It doesn't have to be this difficult.
I always assumed licensed, translated Japanese comics were made by acquiring the digital masters from the Japanese publishing companies and using staff translators, maybe even in collaboration with the original author. I was very wrong.
Tokyopop, a large importer of Japanese comics, has a video explaining their technique. They have a contact in Japan purchase off-the-self tankobon (compiled volumes) and ship them to the states, where they microwave them to loosen the binding, and scan them in. Then they outsource the airbrushing out the text bubbles and the translation to generic hourly translators, and sell them on US bookshelves for whatever outrageous price they get. I literally thought it was a joke. It's hard to believe a company could have the gall to sell such a product as new. The image quality difference between licensed manga and the very same volume that I have directly imported from Japan is glaringly obvious and they use cheap shit paper too.
Moral? Download fan scanslations. Like most pirated goods, not only are they free, they are better.
would anyone scan a magazine?
200 pages per minute, that would be like 3 minutes a book (big one). When these come out to the public, I want one!
Or learn Japanese. It'll be fun.