Create OCR Layer in PDF with Hazel - FREEWARE (solved)

macOS
1

Hi Guys,
I found some topics about Hazel and OCR but it seems like most of you are using PDFpen Pro to get the OCR layer.
I wanted to ask if someone has a free tool which can accomplish the same. I dont need a new PDF editor, because I have PDF expert and it fits for everything I want to do, exept the OCR thing. I already mailed them but as far as i know this feature is not available.

What I want is (maybe a terminal tool - I think there is something available with brew) a tool which can apply the OCR layer in the background.

EDIT: solution in last reply

2 Likes
2

I tried using Tesseract for a workflow when I was trying to replicate Evernote’s ability to scan images for words. I switched to a Devonthink Pro workflow because Evernote got creepy. It worked okay, but it was not the most accurate tool for my task.

1 Like
3

Not free, exactly, but I played around with Prizmo, which is included in SetApp, for this. I did get something working at the time (this would have been about a year ago now) as the app has an Automator action. However it took FOREVER to run and I never took the time to look into it further.

PDFPen is included in SetApp now too, though, which may mean the above is irrelevant—being currently Mac-less I haven’t looked at it again.

2 Likes
4

Take a look at ocrmypdf. I have used it quite a bit for documents I scan with my phone. Free and capable, with good documentation.

I used to use DevonThink, but have moved away from it because this seems to work nearly as well and is more convenient for what I do.

1 Like
5

For anyone interested,
thanks to @tbrown313, I installed OCRmyPDF with homebrew (HERE is the instruction).
to automatically scan the incoming PDFs I created the following hazel rule:

11
Bildschirmfoto 2019-05-08 um 19.05.11.jpg1504Ă—646 111 KB

the embedded script is quite simple:

ocrmypdf -l deu+eng "$1" "$1" --skip-text

the --skip-text jumps over every page which already contains text (scans from my iPhone already contain a OCR layer).
you can change the languages (in my case German and English) according to your personal needs.

7 Likes
6

Is anyone still using this? I can’t seem to get the embedded script FlohGro typed up to work. The log doesn’t give me anything useful: Shellscript exited with non-successful status code: 127.

Anyone have an idea for me?

7

Did you install OCRmyPDF?

8

Yes! It works great as a command in the terminal, just not as a script in Hazel. It seems like (the error) has something to do with the file referencing “$1” “$1.” Is it still working for you?

9

Ok… unfortunately I’m not using the script anymore (I let DEVONthink OCR everything)

10

Why is “$1” there twice? Also, shouldn’t there be a named output file at the end, after “skip-text”?

11

If you want to add a OCR layer the command:

ocrmypdf Input.pdf Output.pdf

is referenced in the „cookbook“ of ocrmypdf

That’s exactly what the double „$1“ is doing - in my usecase I didn’t want to receive another file and instead overwrite the current one.

You can of course modify this until it fits your need.

As I said, this worked for me until I switched to DEVONthink

12

Thanks for this script, for use together with Hazel is the best option for me. So I did install OCRmyPDF and tesseract as well. But I get anytime an error in the Skripteditor for the part “l deu” (for me as well german and english are relevant, most german).

What could be the problem here? I’m at a loss.

13

Can you post the error? Did you copy the command from above including the dash „-“?