Question for the book scanners

[Eric Rechlin]

I've been scanning a lot of books for my new site lately, and everything has been possible with free tools.

I use ImageMagick to cut two pages into one. I script it so I don't have to manually do it for each file, but here are sample commands for one file:

convert -crop 1656x2540+5+0 hp42s-prog-de-000.png out/hp42s-prog-de-000a.png;
convert -crop 1656x2540+1656+0 hp42s-prog-de-000.png out/hp42s-prog-de-000b.png

That starts at position 5,0 and grabs 1656x2540 for the left page, and starts at position 1656,0 and grabs 1656x2540 for the right page, and saves them in files with "a" and "b" appended to the original filenames.

I then do a mix of automatic and manual cleanup of the pages, optimize them with optipng and some tools I wrote myself, and then run through tesseract-ocr (a modified build I made to fix bugs) to convert each page to PDF, and then the Popper PDF tools to combine the pages into a single PDF.