Slashdot Mirror
The DIY Book Scanner
azoblue writes "Daniel Reetz did not want to lug around heavy textbooks, so he built a book scanner to create digital copies. '... over three days, and for about $300, he lashed together two lights, two Canon Powershot A590 cameras, a few pieces of acrylic and some chunks of wood to create a book scanner that's fast enough to scan a 400-page book in about 20 minutes (PDF). To use it, he simply loads in a book and presses a button, then turns the page and presses the button again. Each press of the button captures two pages, and when he's done, software on Reetz's computer converts the book into a PDF file. The Reetz DIY book scanner isn't automated — you still need to stand by it to turn the pages. But it's fast and inexpensive.'"
Except for the lack of an automatic page-turner, Daniel's device is the same as one you can buy commercially for about $20,000 (http://www.treventus.com/bookscanner_pageturner.html).
He was wise to decide on manual page-turning.
I'm a Programmer. That's one level above Software Engineer and one level below Engineer.
http://bkrpr.org/doku.php
Same thing, much cheaper (I built mine for ~150 USD.)
http://CryoLANparty.com/ A lan I'm staff on!
You must not have ever gone to college. A textbook for $15? Get real.
You haven't actually tried this have you? I've had various flatbed A4 scanners over the years, all at much higher resolution than a camera, and hence all got down-sampled afterwards for my display that is only 1.5MP anyway. Then I switched to using a phone camera with only a 2MP CCD, but a really good lens and decent macro mode (Sony-Ericcson Cybershot for those that are interested). As long as the focus was good it produced perfectly readable shots, and so it became my portable scanner. These days I mostly shot stuff at home so I have a 12MP DSLR to hand. It's huge overkill, and I massively down-sample stuff afterwards, but entirely readable. So your basic claim that this can't be done with a camera based on the resolution compared to a scanner is a complete load of bollocks. The focus of the lens tends to be the important issue.
Slashdot: where don knuth is an idiot because he cant grasp the awesome power of php
from the comments with the article
posted by: irrational | 12/11/09 | 11:56 pm
I do it in 5 steps, and you get rid of the book when you’re done since you don’t need to store it. After you get done putting 200 hours into your creation, you’ll have spent thousands of dollars worth of your time. I solved this problem much more quickly years ago:
1. Buy a good sheet-fed and high-speed scanner. I have a Panasonic KV-S2026 color.
2. Get a decent jigsaw from Home Depot. Use metal cutting blades (24 teeth/inch or better)
3. Saw the spines off the book and for God’s sake use some C-clamps on each end of the book. Preferably sandwich them between two flat boards.
4. Remove and feed sheets through the scanner to OmniPage and text recognize the pages.
5. Save as PDF.
6. Repeat. You now have searchable digital books!
I'm amazed at how good OCR has gotten. I did the same thing without building anything: just connected my Canon PowerShot A540 to a tripod, lay the tripod on a coffee table, put the book on the floor, and started snapping away. Fed the JPGs to ABBYY FineReader 10, and it spit out plain text that was *at least* 97-98% accurate on every page. I did not use any special lights, do not know anything about photography, and frankly thought I'd have to buy all sorts of special equipment. The only other thing I added for convenience sake was Dirk's CanoRemote so that I would not move the camera (however imperceptibly) every time I pressed the shutter.
The school is NDSU. Yes we (he) looked. No our library does not have one.
He has details of the reasons on his blog danreetz.com/blog
Actually, the motivation behind the project stem's from Dan's stay in Russia before his graduate studies. He realized that their are tons of old posters, pictures, and other soviet propoganda floating around the country's libraries that many people in the western world would like to view, but are unwilling to go to Russia to see. He wanted to digitize some of these posters (works of art, in his view) in order to circulate them on the web. He soon became very frustrated with using a flatbed scanner, and stopped. Zoom ahead a few years later, Instrucatables is having a contest to win an epilog laser cutter, so he decided to build a book scanner out of recycled (read: trash) materials and submits the project, and wins. He says he's surprised at how well the project has resonated with the web community.
http://www.geocities.jp/takascience/lego/fabs_en.html
turning the pages and scanning is childs play