Slashdot Mirror


Scan a Book In Five Minutes With a $199 Scanner? (teleread.com)

New submitter David Rothman writes: Scan a 300-page book in just five minutes or so? For a mere $199 and shipping — the current price on Indiegogo — a Chinese company says you can buy a device to do just that. And a related video is most convincing. The Czur scanner from CzurTek uses a speedy 32-bit MIPS CPU and fast software for scanning and correction. It comes with a foot pedal and even offers WiFi support. Create a book cloud for your DIY digital library? Imagine the possibilities for Project Gutenberg-style efforts, schools, libraries and the print-challenged as well as for booklovers eager to digitize their paper libraries for convenient reading on cellphones, e-readers and tablets. Even at the $400 expected retail price, this could be quite a bargain if the claims are true. I myself have ordered one at the $199 price.

3 of 221 comments (clear)

  1. The actual big news here: by tlambert · · Score: 4, Interesting

    The actual big news here: The company doing the indiegogo is located in Shenzhen, China.

    This is the first one of these I've seen. It struck me as very odd that the video narrator was an almost perfect midwest accent, but had terrible grammar and word choice, but when looking at the location of the startup, it became more obvious that this was actually an Indiegogo out of China.

    Anyway, good on them; I expect that we will be seeing a lot more people doing crowd-sourcing from non-U.S. locations, given that VC thends to be pretty tight outside of specific regions of the U.S. (which is, in turn, why most startups that go anywhere are U.S. based, rather than being in Europe, or elsewhere, where the funding climate is pretty terrible).

  2. OCR is the main problem by DrXym · · Score: 4, Interesting
    I read a lot of books from OpenLibrary (an awesome resource for old books). Most e-books are offered for download in EPUB and PDF format. The PDF is a direct book scan, the EPUB is OCR'd from the scan. Invariably the EPUB is filled with errors caused by OCR - hyphenated words not joined back together, page numbers appearing in the middle of text, words autocorrected to something else, chapter headings screwed up etc. Sometimes the OCR gives up entirely.

    It's simply easier to read the PDF although the file size is enormous and you're basically looking at images of some yellowing old book which means lots of panning and zooming particularly on small devices. And forget reading it on an e-reader.

    So yeah I think you could automate scanning of books, but the second step of getting it into EPUB format is the tricky part.

  3. Re:Welcome to 2006 by Panoptes · · Score: 3, Interesting

    There are two curses of modern book publishing that cause problems whatever hardware you use. The first is so-called 'perfect binding' in which the folds of page gatherings, through which the sections are traditionally sewn together, are instead sliced off and glued to make a rigid spine with an exceedingly narrow angle of opening; the second is the use of low-grade, thin paper with high show-through that mucks up the scan.

    The best software I've found to scan and collate is Softi ScanWiz. With it you may scan one stack of pages, flip the stack and scan the other side - the program then shuffles the page images into the correct order. It also automatically adjusts brightness and contrast so as to minimise ink show through.