Face Blur

Face Blur — Anonymise People in Photos Before Sharing

Not every person in a photo wants to be identified. Children at a school event, bystanders in a street photo, participants in a research study, colleagues in a workplace image you're posting publicly — sometimes you need to share the image but protect the identities of the people in it. Blurring faces is the standard way to do this, and this tool makes it straightforward.

Upload a photo, and the tool automatically detects faces in the image. Detected faces are highlighted so you can confirm what's been found. Then blur them all at once or selectively, choose the blur intensity, and download the anonymised image. Everything runs in your browser — the photo doesn't go anywhere.

How to Use It

Upload your image. The face detection model analyses the photo and draws boxes around detected faces. Review the detections — the tool should catch most clearly visible front-facing faces, though profile views or partially obscured faces may not always be detected. Apply blur to all detected faces at once, or toggle individual face detections on and off to blur only specific people. Adjust the blur strength — a subtle blur may still allow recognition; use a strong blur or pixelation effect to ensure anonymity. Download when done.

If the automatic detection misses a face, most versions of the tool allow you to draw a manual blur region over any part of the image. This is useful for blurring non-face identifying features — visible tattoos, vehicle number plates, street addresses, or documents visible in the background of a photo.

Common Use Cases

Protecting children's identities: Schools, sports clubs, and parents often want to share event photos online without exposing children's faces publicly. Blurring children before posting to Facebook, a school newsletter, or a public website is good practice — and increasingly expected under privacy norms.

Street photography and journalism: Street photographers who want to share work publicly sometimes blur faces of people who didn't explicitly consent to being photographed. This is a common practice in documentary photography shared on social media or portfolios.

Research and academic use: Research papers, case studies, and presentations that include photos of research participants or patients need to protect identities. Blurring faces in all shared images is standard ethical practice.

Corporate and workplace photos: HR teams and internal communications often need to share workplace photos without exposing individuals who haven't consented to being featured. Events, office photos, and team pictures sometimes need selective blurring before going into company newsletters or press releases.

User-generated content moderation: Community managers dealing with reported photos may need to redact identifying information before sharing in escalation reports or documentation.

Number plate blurring: If you're posting a photo of a vehicle — selling a car, showing a parking situation, documenting a road incident — you may want to blur the number plate to protect the owner's privacy. Use the manual region selection to cover number plates that aren't automatically detected as faces.

Tips for Effective Anonymisation

Blur intensity matters. A light Gaussian blur may look obscured in a small preview but can sometimes be partially reversed by image processing techniques. For strong anonymisation, use a heavy pixelation effect rather than a light blur. Pixelation is harder to reverse than smooth blurring.

Faces aren't the only identifying features. Distinctive tattoos, clothing with a name on it, a visible ID badge, recognisable jewellery — these can identify people even when the face is blurred. If complete anonymity is required, use the manual blur region to cover anything that might identify a specific person.

Check the image at full resolution after blurring. What looks thoroughly obscured in a small preview sometimes shows more detail at full size, especially with lighter blur settings.

Privacy of the Tool Itself

Face detection and blurring run entirely in your browser using a JavaScript model. The images you upload are never sent to a server, stored anywhere, or used for training or analysis. This is important for a tool that processes photos of people — the tool is designed so that you never have to hand your images to a third party to use it.

Limitations

Automatic face detection works best on clear, well-lit, front-facing faces at a reasonable size in the frame. Faces in profile, faces partially obscured by sunglasses or masks, small faces in the background of a crowd, and faces in poor lighting may not always be detected. For missed faces, use the manual selection tool. The tool processes individual photos — batch processing of multiple images isn't currently supported.