Issue with two column format

[werkstattcodes]

Hello,

many thanks for all your efforts put into this project.

I am currently working on session transcripts which are in two column format. I had previously tried tesseract (bundled in R package tesseract and pdf_tools) but my results were not totally satisfying.

I now tried your approach, but to my surprise, it only recognized the header line of the page. Any suggestion what I am missing?
Attached I am sending you a sample page of the records. If your time permits, any help/suggestion would be very welcome.

Many thanks

imfname_158653_page4.pdf

[gabriel-piles]

Thank you for reaching out.

The issue with the PDF is that some text is marked as hidden, making it unavailable in the XML format after conversion using Poppler's PDF-to-XML tool.

Furthermore, because the two-column layout is not properly detected, the text could be difficult to understand.

If you want to make it work, clone this repository:

git clone https://github.com/huridocs/pdf-document-layout-analysis
cd pdf-document-layout-analysis

Move to the following branch:

git checkout xml-hidden-by-default

And then execute the text extraction:

make start
curl -X POST -F 'file=@/PATH/TO/PDF/pdf_name.pdf' localhost:5060

Let us know how it went.
Best.