Skip to content
Blur Document

How to redact a scanned PDF or a photo of a document

Scanned PDFs and phone photos of paper have no text to select. How OCR finds names and numbers anyway, what to check by hand, and how to export a safe copy.

Updated · 2 min read

Do it now: free, in your browser, nothing uploaded.

Open Redact PDF

The short answer

Open the scan or photo in Redact PDF. The page is read with OCR on your device, so account numbers, ID numbers, emails and other details are highlighted even though the file is just an image. Check the page by eye, draw boxes over anything missed, and export.

Why scans are different

A PDF created from a word processor contains real text. A scanned PDF, or a photo of a paper form, is a picture of the page. You can't select or search it, so tools that rely on the text layer find nothing to redact.

OCR (optical character recognition) reads the words from the image. Blur Document runs OCR in the browser, and it copes with pages that were scanned sideways or upside down, which is common when a stack of paper goes through a scanner or a phone photo is taken at an angle.

OCR isn't perfect. Faint print, stamps over text, creases and handwriting can all be misread or skipped, so treat the automatic highlights as a starting point, not a guarantee.

Step by step

  1. Drop the scanned PDF, or a JPG, PNG or WebP photo, into Redact PDF. Photos also work in Blur text in image.
  2. Wait for each page to be read. Detected details are highlighted.
  3. Zoom in and read the page yourself. Look especially at handwritten entries, signatures and stamps, which aren't found automatically.
  4. Draw boxes over anything missed. For a QR code, check that it's covered too; detected codes are highlighted.
  5. Export. The output is OCR-read again and anything still legible is flagged, which helps with small numbers you might have missed.

Getting a better scan

  • Lay the paper flat in even light, without shadows across the text.
  • Fill the frame with the page; crop away the table or background.
  • Use a higher resolution if your scanner offers it. Tiny print is the first thing OCR misses.
  • Photograph each page separately and open them together in batch mode, up to 20 files at once.

Photos taken on a phone can carry GPS location data. The exported copy has EXIF data, including location, removed.

For a document like a cheque or ID card, a ready-made template is faster; see what to hide on a cancelled cheque or sharing ID documents safely.

Questions

Can I redact handwriting in a scanned document?

Yes, by drawing a box over it. Handwriting isn't detected automatically, so check every page for it. Signatures next to a printed Signature or Signatory label can be covered with the Cover signatures button.

Does the scan get uploaded for OCR?

No. OCR runs in your browser, and your document stays on your device.

Tools from this guide

More guides