Text extraction from a Image using OCR stored in PDF

[Prasaderp]

So, I have a PDF which has a Title stored using a Image and also normal text. How can i extract both normal text and OCR text on the image together using PyMuPDF. I am able to extract text from PDF but i also want to extract the OCR image text too from same PDF which actually is the Title of the PDF.

Seehere in the Image below I can extract all the normal text which I have selected using ctrl+A but here u can see some Text inside the Images eg: Attention to, Farm etc caanot be extracted. How can I achieve that too!

[JorjMcKie]

PyMuPDF supports "partial OCR":

• accepts "normal" / standard PDF text
• adds text found in images

... get_textpage_ocr which results in a joint "corpus" of all text on the page. The extraction sequence of this is however

1. standard text
2. OCR-ed text

So you need to use sorting by geometrical information when required. A good first approximation can be achieved by this snippet

textpage = page.get_text_ocr(dpi=150, partial=True,...)
blocks = page.get_text("blocks", textpage=textpage, sort=True)

[Prasaderp]

Thanks! This helps a lot.