How to Censor Photos Online - Blur Faces, Pixelate & Redact Images

The Short Answer

Open the free Censor Photo tool and drop in your picture.

Click Detect faces or Find sensitive text, check the boxes it draws, pick blur, pixelate or a black box, and save.

Everything runs in your browser, so the photo never leaves your device. The saved copy has no hidden layers and no EXIF or GPS data. The rest of this guide covers which method to pick, what people forget to hide, and the mistakes that let censored details leak back out.

Blur, Pixelate or Black Box?

All three hide things, but not equally well.

We ran the same made-up card number, email and phone number through each one:

A light blur only softens the edges - your brain fills in the rest. Researchers have also shown that weak pixelation of text can be reversed, by matching the blocks against a known font (the open-source Depix tool does exactly that). So the rule is simple:

How to Censor a Photo Online

1. Open Censor Photo and drop in a JPG, PNG, WebP, GIF, BMP or AVIF (up to 40 MB).
2. Click Detect faces to box every face, or Find sensitive text to find emails, phone numbers and card numbers. You can also draw areas yourself.
3. Pick the shape: box, oval (good for faces) or freehand brush.
4. Pick the effect: blur, pixelate, black box or a colour box, and set the strength.
5. Zoom in and check every area - move, resize or delete boxes as needed.
6. Save as PNG, JPG or WebP.

Want to hide everything except the subject? Use Blur background with a rectangle or oval focus, or Blur entire image for a teaser or a spoiler.

Let the Tool Find Faces and Text

Finding every face in a crowd by hand is slow, and it is easy to miss one.

Faces

Detect faces has two modes. Fast loads a tiny model (about 190 KB) and handles normal photos. Thorough loads a bigger one (about 5.6 MB) and catches small, turned or partly hidden faces in group shots.

Each face gets a box you can review: untick false alarms, drag or resize a box, then press Apply. Add some padding so hair, ears and jawlines are covered too - a recognisable outline can be enough to identify someone.

Sensitive Text

Find sensitive text reads the picture in your browser and boxes:

• Email addresses, phone numbers and links
• Card numbers (checked with the Luhn test, so random digits are skipped)
• IBANs and IP addresses
• Any custom words you type in - a name, a street, a company
• Optionally dates and long ID numbers

It is a strong helper, not a guarantee. Handwriting, tiny print and text at an angle can slip through, so look over the picture yourself before you save.

What to Hide Before You Share

Faces are the obvious part. These are the details people forget:

Children, crowds and anyone who did not agree to be in the photo deserve a box too. In the EU and UK, a clearly identifiable face is personal data, so blurring strangers is the safe default when you publish for a business.

Hidden Data: EXIF and GPS

A photo can give away your location even when nothing in the picture does.

Most phones save the GPS position, the date and the camera model inside the file as EXIF data. Some social networks strip it; messaging apps, email and file-sharing links often keep it.

The Censor Photo tool re-encodes the picture on export, so the saved file has no EXIF or GPS data. For a photo you do not need to censor, use the EXIF Data Remover. To see what a file is carrying first, open it in the EXIF File Viewer.

Censor Many Photos at Once

Got a whole event folder or a batch of screenshots?

Add several images, switch the scope to All images, and run face or text detection on the whole batch. Check each picture, then click Download all - more than two images come as one ZIP file.

To make the batch lighter before you post it, run it through the Image Resizer or Compress JPG afterwards.

Screenshots, Documents and PDFs

For text in a picture, always use a solid box.

For a PDF, censoring a screenshot of it is not enough if you also share the PDF. Drawing a black shape over text in a PDF editor often leaves the text underneath, where anyone can copy it out. Use Redact PDF, which removes the text itself, not just covers it.

Want to point at something instead of hiding it? The Image Annotator adds arrows, boxes and labels to screenshots - and works well together with censoring.

Common Mistakes

1. A Light Blur on Text

It looks hidden on your screen, but zoom out and it is often readable. Use a black box for any text.

2. Cropping Instead of Censoring

In 2023, bugs in the Google Pixel Markup editor and the Windows Snipping Tool ("aCropalypse") left the cut-off part of screenshots inside the saved file, where it could be recovered. Phone editors also keep the original so you can undo an edit. Cropping with a tool that writes a new file, like the Image Cropper, avoids both problems.

3. A Semi-Transparent Highlighter

A marker stroke at 80% opacity can be lifted with a simple levels or contrast change. Use a fully opaque box.

4. Hiding the Face, Not the Name

A blurred face next to a name tag, a username or a distinctive tattoo still identifies the person.

5. Saving in a Layered Format

A PSD, a PDF with annotations, or an editor's project file keeps the original under the censor layer. Export a flat PNG, JPG or WebP.

6. Sharing the Original Too

Check that the uncensored version is not also in the same album, chat or cloud folder link.

Frequently Asked Questions

The Bottom Line

• Text? Use a solid box.
• Faces and plates? Strong pixelate or strong blur - never light.
• Look beyond faces: reflections, screens, barcodes, name tags
• Save a flat new file - no layers, no EXIF, no undo history

Still having trouble? Please contact us and we'll be happy to help!