1000 pages per second?! It captures the pages AS they're flipped.
http://www.engadget.com/2009/08/13/book ... urbo-mode/
Fastest Book Scanner Around
Moderator: peterZ
-
spamsickle
- Posts: 596
- Joined: 06 Jun 2009, 23:57
Re: Fastest Book Scanner Around
I think they're saying 1000 frames per second, not 1000 pages per second. I doubt that you could actually flip 1000 pages a second and get usable images.
They have to be doing some heavy-duty image processing on the back end. I'm not going to believe it doesn't come out looking like crap until I see examples of the final product. I know how hard it is to get good results with flat pages that are turned one at a time.
They have to be doing some heavy-duty image processing on the back end. I'm not going to believe it doesn't come out looking like crap until I see examples of the final product. I know how hard it is to get good results with flat pages that are turned one at a time.
Re: Fastest Book Scanner Around
This looks like something that has been done before. Google has patented a system which analyze the curvature of the pages in order to process the data from them. Google most likely has (at their disposal) a few super computer and high-def high speed cameras. So people will figure that's what they've been doing all this time - scanning at spellbound speeds. Yet their patent doesn't seem to mention how they process their data. Their spokesperson are very secretive about how many pages per minute they can capture. I first heard about the story at http://www.npr.org/blogs/library/2009/0 ... 08978.html
If the university wants to expand on their technique they will have to come face to face with Google's patent. And by the looks of it Google wants to be the first to scan out of print books before someone else gets their hands on them. And this machine probably scares the living bajeesus out of them. I believe (Daniel and other's) affordable DYI book scanners are helping by making sure people archive their books (especially to be out-of-print books) before it's too late.
Google didn't invest in a page turning mechanism probably because it only takes a few seconds to stand in front of the camera to flip the pages. They'd figure computer muscles are cheaper than a clunky robotic arm. If two pages are stuck together, for example during page turning, just flip the pages again and the computer will know how to weed out good pages.
It's possible to capture a 1000 page book within a second. You might need to punch holes in the pages so that the air resistance won't slow down larger pages. Or your book has to be super small and stiff like a spring. But the point is that this machine will be a formidable foe to Google's already secretive super fast scanners.
(the image above is captured/merged from a video source)
If the university wants to expand on their technique they will have to come face to face with Google's patent. And by the looks of it Google wants to be the first to scan out of print books before someone else gets their hands on them. And this machine probably scares the living bajeesus out of them. I believe (Daniel and other's) affordable DYI book scanners are helping by making sure people archive their books (especially to be out-of-print books) before it's too late.
Google didn't invest in a page turning mechanism probably because it only takes a few seconds to stand in front of the camera to flip the pages. They'd figure computer muscles are cheaper than a clunky robotic arm. If two pages are stuck together, for example during page turning, just flip the pages again and the computer will know how to weed out good pages.
It's possible to capture a 1000 page book within a second. You might need to punch holes in the pages so that the air resistance won't slow down larger pages. Or your book has to be super small and stiff like a spring. But the point is that this machine will be a formidable foe to Google's already secretive super fast scanners.
- TokyoUniversityScan.jpg (73 KiB) Viewed 5670 times