CanDoYa
All tools
Saved toolsYour favorites

Add to favorites

Saved in this browser only.
Private browser utility / Images

Face Blur Online

Runs entirely in your browser - no upload, no sign-up.

Live workspaceLocal processing

Choose a photo to find and anonymize faces.

Share this tool
face blur anonymizer / browser utility
01 / Overview

What is a face blur tool?

A face blur tool finds faces in a photo and covers identity-bearing details before you share it. This face anonymizer runs MediaPipe detection in your browser, lets you review every region, and exports a protected PNG. Your source photo and finished image are not uploaded to CanDoYa.

02

How to use

  1. 01
    Choose a photo

    Drop in a JPEG, PNG, or WebP up to 20 MB and 16 megapixels, or load the example photo.

  2. 02
    Detect and review faces

    Run the on-device detector, check every marked area, and draw a manual box around any face it missed.

  3. 03
    Choose protection

    Select blur, pixelation, or blackout, then adjust effect strength and extra face coverage while watching the preview.

  4. 04
    Download the result

    Review the entire image at full size and save the protected PNG to your device.

03

Who it's for

  • Parents and schools preparing event photos while protecting children and bystanders.
  • Journalists and researchers anonymizing interview, protest, or fieldwork photos before distribution.
  • Workplace teams redacting faces from incident reports, presentations, training material, and internal screenshots.
  • Social media users hiding strangers or friends who did not consent to appear in a public post.

Automatic detection is a starting point, not a privacy guarantee. The editor uses MediaPipe's full-range BlazeFace model, marks each detected face, and lets you disable a region, add a missed face manually, or expand coverage. Processing and PNG encoding happen on your device.

Soft blur preserves the scene's look, while coarse pixelation makes the edit obvious. A solid blackout removes the most visual detail. The output is always PNG so repeated edits do not add JPEG compression around protected areas.

FAQ

Is my photo uploaded to a server?

No. The selected photo is decoded, detected, edited, and encoded inside your browser. The worker downloads the public MediaPipe runtime and BlazeFace model from external hosts, but those requests do not contain your image. CanDoYa does not receive or store the source or result.

Is the face blur tool free?

Yes. There is no account, payment, credit limit, or watermark. Your device supplies the computing power. The first detection downloads public model and runtime files, so it needs an internet connection and may use data on a metered network.

What image limits apply?

The tool accepts one JPEG, PNG, or WebP image up to 20 MB and 16 megapixels, with each side at least 64 pixels. These limits reduce large browser memory spikes. The downloaded PNG keeps the source pixel dimensions and may be larger than a compressed JPEG.

Will automatic face detection find everyone?

No detector is perfect. Profiles, masks, very small faces, motion blur, harsh shadows, and partial occlusion can cause misses. Review the full image before download and use Add a missed face to draw a region around anything the detector did not protect.

Is blurring a face enough to anonymize someone?

Not always. A light blur can retain recognizable structure, and people may still be identified by hair, clothing, tattoos, surroundings, or metadata. Increase coverage and strength, or choose coarse pixelation or blackout for sensitive material. Always inspect the exported image before sharing it.

What is the difference between blur, pixelate, and blackout?

Blur softens facial detail and blends naturally into a photo. Pixelation replaces detail with large color blocks, making the edit more obvious. Blackout covers the region with a solid shape and reveals the least facial information, so it is the safer choice when concealment matters most.

Does face detection work without a GPU?

Yes, on browsers that support WebAssembly and module workers. The tool tries MediaPipe's GPU delegate first when available and falls back to CPU processing if setup fails. CPU detection can take longer, but the photo still stays on the device.

Does the downloaded image keep photo metadata?

The tool draws the visible image pixels onto a canvas and creates a new PNG, so original JPEG EXIF fields are not copied into the exported file. Still verify sensitive workflows independently because filenames, surrounding context, and other files can reveal identity.