Image Blur and Pixelate Tool – Hide What Shouldn’t Be Shared
Screenshots and photos often carry more than you meant to show: a face in the background, a licence plate, a name in a chat window, an email address, the last digits of a card. This image blur and pixelate tool lets you draw rectangles or ellipses over those areas and choose Pixelate, Blur or a Solid box for each one. It also works the other way round: mark your subject as Keep sharp and blur, pixelate or fill everything else, which gives you a quick background blur or a spotlight effect. With no shapes at all, a background effect covers the whole picture. Everything runs on a canvas in your browser, so the image is never uploaded.
Choosing an effect
| Effect | Best for | Can it leak? |
|---|---|---|
| Solid box | Text, numbers, card details, QR codes, signatures | No. The pixels are replaced outright. |
| Pixelate | Faces, plates and logos when you want the area to look natural | Yes, when the blocks are small relative to the area. |
| Blur | Faces, backgrounds and soft privacy edges | Yes, when the blur radius is small relative to the area. |
Why pixelated text isn’t safe
Pixelation replaces each block with the average colour of the pixels inside it. That sounds destructive, but the averages still depend on which characters were there. In 2020 the security researcher Sipke Mellema published Depix, an open-source tool that recovers passwords from pixelated screenshots by rendering candidate text in the same font, pixelating it the same way and matching the blocks. Blur leaks in a similar way, and a QR code pixelated at roughly its own module size can still scan. That is why the tool shows a permanent tip recommending a Solid box for any text or numbers, with a one-click switch.
Faces are at risk too. In 2016 Richard McPherson, Reza Shokri and Vitaly Shmatikov showed in Defeating Image Obfuscation with Deep Learning that neural networks can re-identify people from pixelated and blurred photos far more often than humans can. So each Hide shape gets a strength rating based on how many blocks (for pixelate) or blur radii (for blur) fit across its shorter side. Five blocks or fewer, or four blur radii or fewer, rates Strong. More than ten rates Weak, with a warning that features may still be recognisable.
Why the preview can fool you
A 12-megapixel photo is shown on screen at a fraction of its size. Shrinking it averages pixels together, so a weak pixelation can look solid in the editor while the saved file is still readable. The Check at 100 % panel renders the selected area at actual size, side by side with the original, using exactly the code that writes the file. If you can read it there, so can anyone who opens the file. A coverage view also tints exactly which pixels each shape affects, including soft edges.
The file you save is brand new
In 2023 the aCropalypse bugs showed how edited screenshots can betray their originals. Google Pixel’s Markup editor (CVE-2023-21036) and the Windows Snipping Tool (CVE-2023-28303) wrote the edited image over the original file without truncating it, so the tail of the original could be recovered from the “cropped” file. This tool never edits your file in place. The output is built tile by tile on an empty canvas from finished pixels, then encoded as a new -blurred file. Solid boxes are applied before any other effect, so the pixels under them never feed into a nearby blur or pixelation.
Re-encoding also drops EXIF metadata, including GPS location and the embedded thumbnail that some editors leave showing the unedited picture. If you only need to strip metadata without touching the pixels, use the EXIF Metadata Remover instead.
- Cover each area with a margin around it, not just the visible edge.
- Check every shape at 100 % before saving.
- Look for reflections, other screens, name badges and documents in the background.
- Remember that metadata is removed on save, but the original file is not touched.
- Share the new file, not the original.
Limits
There is no automatic face or text detection: a detector that misses one face in a crowd gives false confidence, so you mark every area yourself. Animated images are reduced to their first frame. Files up to 25 MB and 50 megapixels are accepted, and a browser that blocks canvas reads for privacy still offers the Solid box.