How to make a scanned PDF searchable

Why a scanned PDF cannot be searched

A PDF made by scanning paper, or by photographing it with a scanning app, stores each page as an image, the same as a JPG would. A PDF reader can only search for text it actually contains, so pressing Ctrl+F on a scanned document finds nothing, however clearly the words are printed on the page.

What OCR actually changes

Optical character recognition (OCR) reads the letters in that image and places matching, invisible text exactly on top of where each word sits. The page still looks like the original scan, but a reader can now find, select and copy the words, because there is real text there for it to match against.

In Adobe Acrobat

Acrobat Pro's Scan & OCR tool does this well and keeps formatting closely, but it needs a paid Acrobat subscription; the free Acrobat Reader cannot add OCR text to a PDF, only read PDFs that already have it.

On a phone, while scanning

Apps like Adobe Scan and Microsoft Lens can OCR a page as part of scanning it, which is convenient for a fresh document but does nothing for a PDF you already have from somewhere else, such as an old scanned manual or a document someone emailed you.

Making an existing scan searchable without uploading it

Stayput's OCR PDF tool runs Tesseract, an open-source OCR engine, directly in the browser: drop a scanned PDF and it reads the text on your device, adds it invisibly over each page, and gives back a searchable PDF, with an option to also save the text as a plain .txt file. It reads English text; other Latin-alphabet languages are read with the English model and come out less accurately, and it does not yet read Cyrillic, Arabic or Asian scripts. Clean, straight, sharply printed pages read best; faint photocopies and pages photographed at an angle read worse, so scanning them with a document scanner first, which straightens and flattens the page, generally improves the result.

Questions