Slicing and dicing photocopies of open faced books

October 18, 2016 7:33 AM

I'm always copying chapters from books at work. I usually scan them straight to a pdf file by opening the book face down and copying two pages at once. This makes the final product difficult to read on various screens.

I don't know how to copy a single page at a time because I would have to bend back the book so only one page is touching the scanner glass, or copy it upside down, with the other page hanging off the edge of the copier.

Is there some sort of software out there (for any platform) that will look at a two page pdf file, and batch slice and dice it so there is just one page of text per page?

posted by mecran01 to Computers & Internet (13 answers total)

Response by poster: Ok, this is embarrassing, but simply asking the question helped me formulate a Google search that lead to the answer:

Adobe solution

Google search

I would still like to find an open source solution to this problem if anyone knows of one.
posted by mecran01 at 7:39 AM on October 18, 2016 [1 favorite]

You might also poke at your copier- I remember discovering such a 2 page to one page PDF feature deep within the settings of a copier once.
posted by rockindata at 7:56 AM on October 18, 2016 [4 favorites]

I used ScanTailor when I was in college. It's a little fiddly but it works.
posted by Ampersand692 at 8:05 AM on October 18, 2016 [4 favorites]

Seconding ScanTailor for the page split.
posted by Obscure Reference at 8:13 AM on October 18, 2016 [2 favorites]

Sounds like you've got software Solutions from the above comments. I just wanted to share what I used to do at work because we weren't in control of what software we could have installed. I would scan each set of two facing pages twice, then use the Crop tool in Acrobat to cut off half of each page of the resulting PDF.
posted by The Underpants Monster at 8:59 AM on October 18, 2016 [1 favorite]

Response by poster: The Underpants Monster, that is ingenious. These are all just great solutions, and I wish I had asked this question fifteen years ago.

I suppose that really good OCR might solve this problem too, but I am old and don't trust it.
posted by mecran01 at 9:07 AM on October 18, 2016

Alternate solution, which sometimes works better for me with tightly bound journal volumes (i.e. it started as individual journal issues, they've been bound into a larger volume, there is not much margin, and you can lose a lot of text in the curve of the binding.

- Scan the first page whichever way lets the other half the book hang off the edge of the copier (to get as much flat surface as possible.
- Flip book 180 degrees, scan next page so the other half the book hangs off the edge of the copier.
- Repeat for however many pages
- In the resulting PDF, go through and flip pages (depending on software, this is a 1-2 click thing per page, or can sometimes be done automatically on alternate pages)
posted by modernhypatia at 9:47 AM on October 18, 2016 [1 favorite]

Abbyy fine reader does a good job of this as part of its ocr process, but not free, not open source, and definitely ocr.
posted by crazy with stars at 10:06 AM on October 18, 2016 [1 favorite]

I'm not sure how far your patience for "any platform" or your enthusiasm for open source go, but if you have access to (or are willing to install in a virtual machine) a unix-like environment (probably including OS X), and if you're willing to fiddle with a few lines of shell scripting language and run a command in a terminal, you can automate the whole thing pretty easily using ghostscript commands.

If you want to uniquely specify which regions of the scan get turned into the odd and even pages, it will take a bit of fiddling to specify the clipped region for each specific book. If you're content to split the page down the middle every time, then it can be fully automated. Crawl through the document and save the left half of page n to file "na", do it again and save the right half to file "nb", then concatenate them in order to generate a final document and delete the temporary files.

(In the somewhat unlikely event this sounds like something you want to do, I'd be happy to create a working example for you.)
posted by eotvos at 1:13 PM on October 18, 2016 [1 favorite]

unpaper can do this from the command line, but it's a little fiddly.

I don't know if MeFi's Own™ Matt Zucker ever developed Page dewarping beyond a concept.
posted by scruss at 3:01 PM on October 18, 2016 [1 favorite]

On OS X (macOS) this is pretty easy and has a nice GUI, and works for any number of pages.

1. Open the PDF in Preview.app
2. Click the Suitcase-looking icon ( Show Markup Toolbar)
3. Click the second icon from the left (Rectangular Selection)
4. Click & Drag to select page 1.
5. Edit/Copy
6. File/New
7. Paste. You now have page 1 in a document
8. Repeat steps 4-6 for page 2. You now have 2 PDFs each containing the cropped page.

To combine the documents into a single PDF:
1. For both documents, click the View menu and choose "Thumbnails"
2. You can now drag & drop page 1 into page 2's document (or vice versa)

To combine them into a single PDF:
1. From the File menu choose Print and then select "Open as PDF" or "Save to PDF"
posted by soylent00FF00 at 5:29 PM on October 18, 2016 [2 favorites]

It's a little outside of the scope of the question, but the Internet Archive has a V shaped setup for scanning (or photographing) both pages of an open book simultaneously. You turn the page yourself but then they have one button or foot pedal to take both photos at the same time.

Might be worth considering if you have some affordable cameras and a script to download and collate the photos, even if you end up having to take both photos yourself manually since you're probably only doing one chapter instead of the whole book.
posted by cali59 at 6:48 PM on October 18, 2016

Response by poster: Ok, so the easiest method I have found so far for OS X is a free program called Briss written in Java, so it's cross platform.
posted by mecran01 at 7:13 PM on October 27, 2016