How to Make a Scanned PDF Searchable (OCR)
You open a scanned contract, press Ctrl+F to find a name — and nothing is found, even though you can see the name on the page. Copying text selects the whole page as a picture. The reason is simple: a scan is a photograph of paper, not text.
OCR (optical character recognition) fixes this by reading the letters in the image and adding them as real, searchable text. This guide explains how it works and how to get the best results.
Text PDFs and scanned PDFs
A PDF exported from Word or another program contains real text: each letter is stored as a character in a font, so you can search, select and copy it.
A scanned PDF contains one picture per page. To a computer, the letters are just dark pixels. Tools that work with text — search, copy, PDF to Text, Compare PDF — find nothing.
What OCR adds to the file
An OCR engine analyses each page image, finds lines of text and recognises the characters. A searchable PDF then keeps the original page image exactly as it was and places an invisible text layer on top, aligned with the words in the picture.
The page looks identical, but now Ctrl+F finds words, you can select and copy text, and other tools can extract it.
Make a PDF searchable with FilesSwift
The OCR PDF tool uses the open-source Tesseract engine:
- Open OCR PDF and upload a scanned PDF, or a photo of a document (JPG, PNG and similar).
- Press Convert. Each page is rendered at 200 DPI and recognised, so long documents take a little while.
- Download the searchable PDF and test it by searching for a word you can see on the page.
How to get accurate results
OCR accuracy depends almost entirely on the image. Scan at 300 DPI if you can, keep pages straight, use even light without shadows, and avoid motion blur when photographing with a phone.
Clean printed text in common fonts is recognised very well. Handwriting, decorative fonts, very small print, faded photocopies and text on busy backgrounds are recognised poorly.
Recognition in FilesSwift is currently set up for English text. Gujarati, Hindi and other scripts are not recognised yet.
After OCR: what you can do
Search and copy text directly in any PDF reader. Extract all the text with PDF to Text. Compare two scanned versions of a contract with Compare PDF. Or convert the searchable PDF with PDF Converter if you need an editable document — expect to correct some mistakes.
Frequently asked questions
Does OCR change how my document looks?
No. The original page image is kept; the recognised text is added as an invisible layer.
Is OCR 100% accurate?
No. Clean printed text is recognised very well, but always check important details such as numbers and names.
Can OCR read handwriting?
Generally not reliably. OCR engines like Tesseract are designed for printed text.
Can I OCR Gujarati or Hindi documents?
Not with FilesSwift yet — recognition currently supports English text.