Daniel Reetz, the founder of the DIY Book Scanner community, has recently started making videos of prototyping and shop tips. If you are tinkering with a book scanner (or any other project) in your home shop, these tips will come in handy. https://www.youtube.com/channel/UCn0gq8 ... g_8K1nfInQ

Would it be possible a video 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.
Post Reply
geaplanet
Posts: 24
Joined: 20 Jan 2015, 04:17
Number of books owned: 0
Country: spain

Would it be possible a video scanner?

Post by geaplanet » 22 Dec 2019, 06:33

Nowadays there are smartphones that record video in 4k. Would it be possible to record a video by turning the pages of a book to scan a book? I think that taking the precaution of waiting a couple of seconds before turning the next page would be enough.
1- Detect when there is no page movement and extract those frames. With Motion maybe?
2- Of all the extracted frames there are those that are repeated. If we wait 2 seconds on each page pass there will be about 50 repeated pages. Maybe they could be combined to have a better resolution. With Dupeguru maybe?
3- Flatten. Adjust margins or remove fingers. Unpaper, scantailor maybe?
4- Collect and save as pdf or whatever. Many options here.

dpc
Posts: 311
Joined: 01 Apr 2011, 18:05
Number of books owned: 0
Location: Issaquah, WA

Re: Would it be possible a video scanner?

Post by dpc » 24 Dec 2019, 15:23

This idea has been discussed here over the years. You can search through the forums for 'video' and probably find a half-dozen posts about it.

The important point is to ask yourself how much time are you going to save by developing and using this method over the typical dual-camera V-cradle DIY scanner that is popular on this website? A few seconds per page-turn? So instead of spending 15 mins to scan a 300 page book, you'll get it done in 5 minutes, saving 10 minutes of your time per book? How much of that saved 10 minutes is now going to be eaten up by the post-processing of your acquired images that will likely not be near the quality of the images from the dual-camera scanner?

Oh, and then there's your time used to develop the software to extract the sharp images from a video file, remove fingers, compensate for book curvature, etc., when the software is already available for working with captured images from a camera of the typical DIY scanner.

One other point worth mentioning is that you'll need the entire book scanned and the video file written to disk before you can begin to start the post-processing phase. That's not the case when capturing individual images. The post-processing software that I've developed starts working on images right after they are shot. When I get to the last page of the book, it takes less than 15 seconds before the all of the pages are ready to go to Acrobat to create the PDF. They've already been clipped, color-corrected, rotated/straightened, light-corrected, and resized. All this is possible because the images are ready for processing immediately after they have been generated. They're being worked on in the background while I'm still page turning and acquiring subsequent pages. You can't do this sort of optimization if you're capturing a video file.

Anyway, the moral of the story is to ask yourself what you expect to gain from this and realize that in book scanning the amount of work you have to do after acquiring images is almost always more than photographing/turning pages. If your goal is to reduce the time it takes to digitize a book, you're better off spending time improving the post-processing phase, rather than the image acquisition phase. The exception to that rule is if you can do something with your scanner design to improve the quality of the images it produces (flattened pages, remove glare, even lighting, squared images that require no rotation, etc.) you should do it because it can make things easier on the back end when processing the images. The goal should be no manual intervention once the images are acquired.

Happy Holidays!
-d

Post Reply