OCR vector text

[Kevin-carv]

Hello! I have some PDFs that I made notes on using my drawing app. When I open them in Adobe Acrobat, the software recognizes the text as a vector, like a drawing, so it does not support OCR.

So, does OCRmyPDF suports this type of OCR? See this PDF as example! there is vector notes on the last pages

Rufino, 2021.pdf

[jbarlow83]

You can use ocrmypdf with force OCR mode to render everything as an image and add ocr. A drawback is the loss of vector resolution

Ghostscript also has a feature to perform OCR on individual glyphs of each embedded font, which can help when a font is present but lacking a Unicode mapping. This does not work when text is rendered as curves, as it cannot tell the difference between a curve that represents a letter and one that does not.