How to Redact a Scanned PDF Document
To redact a scanned PDF document, you have to change the image itself, because a scan is a picture of a page and a black box drawn on top leaves the original pixels in the file. In Gold PDF, press Ctrl+Shift+O to recognize the text so you can search it, press R to mark each area, and save: every redacted page is rebuilt with the black bars burned into the image.
Quick steps
- Open the scanned PDF in Gold PDF and use Save as to keep a private, unredacted copy.
- Press Ctrl+Shift+O to run OCR (English and Hebrew are bundled) so the page has searchable text.
- Press Ctrl+F and search for each name, number or phrase you need to hide. Note every page it appears on.
- Read every page with your eyes too, looking for handwriting, stamps and anything OCR may have misread.
- Press R and drag a redaction box over each item, a little larger than the text.
- Press Ctrl+S. Gold PDF rebuilds each page that has a redaction.
- Close the file, reopen it, search again and try copying the black areas. Nothing should come back.
Why scanned PDFs are different
A PDF made from a word processor stores real text: letters you can select, copy and search. A scanned PDF usually stores one large image per page, the photograph your scanner or phone took. Sometimes a scanner or an earlier program has also added a hidden OCR text layer on top of that image, so the page looks like a picture but still answers to Ctrl+F.
That gives a scanned page up to two places where the information lives:
| Layer | What it contains | What happens with a drawn black box |
|---|---|---|
| The page image | The pixels of the scanned paper | Untouched underneath; the box can be deleted or the image extracted |
| A hidden OCR text layer (if any) | Machine-read copies of the words | Untouched; the words can still be searched and copied |
| A real redaction in Gold PDF | Both are rewritten on save | The image gets solid black pixels; the removed words are left out of the new text layer |
This is why a quick rectangle or a black highlight is even riskier on a scan than on a normal PDF. You cannot see the hidden text layer, so you may not know it exists, and image extraction tools can pull the untouched page picture straight out of the file. The general problem is explained in does drawing a black box redact a PDF.
Step one: make the scan searchable
You can redact a scan without OCR, by looking at each page and dragging boxes over what you see. For a one-page letter that may be fine. For a 40-page medical file or a bundle of bank statements, it is how things get missed.
Press Ctrl+Shift+O in Gold PDF to recognize the text on the pages. Recognition runs on your own computer; the file is not uploaded anywhere. English and Hebrew language data are included. Once it finishes, Ctrl+F works across the whole document, so you can search for a client's surname, an account number or an address and jump to every hit.
OCR is not perfect, especially on faint, skewed or low-resolution scans. Plan for that:
- Search for parts of a value as well as the whole: the last four digits of an account, a surname without the first name.
- Search for likely misreads: 0 read as O, 1 as l or I, 5 as S, 8 as B.
- Search for the value with and without spaces, dashes and slashes.
- Treat search as a helper, not a guarantee. Every page still gets a visual check.
For more detail on the recognition step, see how to OCR a scanned PDF for free.
Step two: find what search cannot
Scans carry information that OCR rarely reads well. Before you redact, go through the document page by page and look for:
- Handwriting. Signatures, initials, notes in the margin and filled-in boxes on paper forms.
- Stamps and seals. Received stamps often include dates, names and reference numbers.
- Photos and ID cards. A scanned passport or license copy contains a face, a number and often a machine-readable strip.
- Barcodes and QR codes. These can encode an account, case or patient number. If you do not know what is in one, cover the whole code.
- Page edges. Fax headers, scanner footers and fold marks sometimes show a phone number or name in tiny print.
- Back sides and blank-looking pages. Faint text bleeding through from the reverse side can still be readable when someone adjusts the contrast.
Zoom to 200% or more for this pass. Small print on a scan is much easier to miss than on a digital page.
Step three: redact and save
Press R for the Redact tool and drag a box over each item. On scans, draw the box generously: a few extra pixels on every side, because scanned letters have soft, blurry edges and a tight box can leave the tops of tall letters or the tails of digits visible. If a value sits inside a table line or a form box, it is usually safer to cover the whole cell.
When you press Ctrl+S, Gold PDF rebuilds each page that contains a redaction. The page is rendered as a high-resolution image with the black bars burned in, so the original pixels under each bar are replaced, not hidden. The words that remain on the page are then added back as an invisible text layer, which keeps the page searchable. The words you redacted are not in that layer. Pages without redactions are left as they were.
How to check the redacted scan
Never judge a redaction by how it looks on screen. Close the saved file, reopen it, and test:
- Search for each redacted value, its parts and its likely misreads. You should get no results inside the black areas.
- Copy by dragging across a black bar and pasting into Notepad. Nothing from under the bar should appear.
- Zoom in on each bar at high magnification to confirm no edges of letters peek out.
- Check the properties of the file for a title, author or subject that contains a name.
A full checklist is in how to check if a PDF redaction is secure.
Common scanned documents and what to remove
| Document | Usually redacted | Easy to miss |
|---|---|---|
| Bank statement | Account number, address | Account number in the page footer, barcode on the first page |
| Medical record | Patient name, date of birth, record number | Handwritten notes, stamps from the clinic |
| Signed contract | Personal address, ID numbers | Initials on every page, a scanned ID attached at the end |
| Pay slip | Employee ID, tax number, salary | Employer reference numbers in small print |
| Government form | Identity numbers | The same number repeated in a box at the top of each page |
If one of those numbers is a Social Security number, the extra search formats in how to black out a Social Security number in a PDF apply to scans as well.
What Gold PDF does and does not do
Gold PDF is a free, offline PDF editor for Windows. It can run OCR on scanned pages (Ctrl+Shift+O, with English and Hebrew bundled) so you can search them, and its Redact tool (R) replaces the content under each box when you save by rebuilding only the redacted pages. Your files stay on your computer throughout. It does not find sensitive information for you: there is no automatic detection of names, numbers or faces, so locating every item is your job, using search and a careful read of each page. OCR quality depends on the scan, so a poor scan needs a closer visual check. It is not available for Mac, Linux or phones. For a short overview of the tool, see Redact a PDF.
A routine for every scanned file
Save as, OCR, search every value and its variations, read every page at high zoom, redact with generous boxes, save, reopen and test. The extra minutes go into the visual pass, because on a scan the thing that leaks is usually the thing the computer could not read: a handwritten note, a stamp or a code strip in the corner.
Frequently asked questions
Can I redact a scanned PDF if I cannot select the text?
Yes. A scan is an image, so you can still drag a redaction box over any area. Running OCR first only helps you search for every copy; the redaction itself works on the pixels.
Does a black box on a scanned page remove the information?
Only if the tool rewrites the image. A drawn box sits on top of the picture, and the original pixels stay in the file underneath. Real redaction replaces those pixels when you save.
Why does OCR matter before redacting a scan?
Without OCR, a scanned page has no searchable text, so Ctrl+F finds nothing and you have to rely on your eyes. OCR lets you search for names and numbers across many pages.
Will the redacted scan still be searchable?
In Gold PDF, yes. Each redacted page is rebuilt as an image with the black bars burned in, and the remaining words are added back as an invisible text layer, without the words you removed.
Gold PDF — the free PDF editor for Windows
Edit text, redact for real, OCR scans, fill forms, sign and protect. Offline, no account, no watermark.