How To Get Text from an Image Only PDF File Using Linux

This article is for technically savvy people, and especially for those who use Linux on their Desktop or Laptop PC as I do.

I wanted to post a 19th-century book, The Black Pope, but the text from the PDF file I downloaded is not extractable using copy and paste. I knew I needed to use OCR software to get the text. On your PC you need pdftoppm and tesseract installed. pdftoppm was already installed on my laptop by default. I just needed to install tesseract which is OCR software. I use Fedora Linux, but this proceedure will work in any distribution of Linux.

First I made a folder for Black_Pope.pdf and moved to file into the folder. Then I opened Terminal inside the folder.

These are the commands I used in Terminal to get the ASCII text.

pdftoppm -png Black_Pope.pdf black-pope

(This made PNG files of each page of the PDF. There were 404 in all.)

for i in black-pope-???.png; do tesseract "$i" "text-$i" -l eng; done;

(This command scans each of the PNG files and creates .txt files of each of them.

cat text-black-pope* > black-pope-complete.txt

(This combined all of the 404 text files into a single file.)

It only took a few minutes! The PC did all the work. Just think how long it would have taken me if I scanned each one of those PNG files one by one, and combined them all one by one. Probably an hour or more.

I wound up with a single file to proofread. After I proofread a section, I removed all the extra line returns with an online tool: Remove line breaks with paragraph restoration before I copied the text into the WordPress post.