Blur Part of an Image — Hide Faces, Plates and Details
Drag a box over anything in a photo or screenshot you do not want visible — a face, a number plate, an address bar, an account number, a name in a chat log — and blur it, pixelate it, or black it out. The pixels underneath are genuinely replaced before the file is written, so there is nothing left in the image to recover. It all happens on your device.
The reason to care about that last sentence is that the most common way people censor a screenshot does not work. A black rectangle drawn in a document, a slide, or a PDF sits on top of the original, and the original is still in the file — which is how redactions have been undone in published court filings and press releases more than once. Exporting a flattened bitmap, as this page does, removes the possibility entirely.
Blur, pixelate, or black box
A black box is the only one of the three that is unarguable: nothing survives it and the reader can see that something was removed. Pixelation with large blocks is the usual choice for faces and plates because it looks less aggressive while still destroying detail. Blur is the weakest — a light blur over text can sometimes be read, and over a small set of possible values, such as a six-digit code, a reversible transform can in principle be attacked. For anything genuinely sensitive, use a black box or push the strength up until no structure remains.
Cover more than you think you need
People are identifiable from surprisingly little. A blurred face above a visible name badge is not anonymous; a redacted account number next to a visible sort code and a full name is not much better. Look at the whole frame before exporting: reflections in windows and screens, a delivery label in the background, a browser tab title, a street sign, the top of a passport in a pile of paper. Extend each box past the edge of what you are hiding rather than tracing it exactly.
Metadata is a separate problem
Censoring pixels does nothing about the data attached to the file. A photo from a phone can carry GPS coordinates, the device model, and the exact time it was taken, and that survives every visual edit. The export here writes a fresh bitmap from a canvas, which drops the original EXIF block, but if you are publishing photos regularly it is worth checking with Remove EXIF rather than assuming.
Related: black out text in a document with Redact PDF, replace personal data in text with the text anonymizer, or blur a whole background with Blur Background.
How to use Blur Part of an Image
- Drop in a photo or a screenshot. It is loaded into the page — nothing is uploaded.
- Pick the effect for the next box: pixelate for faces and plates, black box for anything that must be unrecoverable, blur for a softer look.
- Drag a rectangle over the area. It is applied immediately so you can see the result at full resolution.
- Add as many boxes as you need — each keeps the effect and strength that were set when you drew it, so you can mix them in one image.
- Check the whole frame for anything you missed, then download. The saved file contains only the censored pixels.
Features
- Three effects in one image — each box keeps the effect and strength it was drawn with, so a black box and a pixelated face can coexist.
- The pixels are replaced, not covered — the exported file has no original underneath it to recover.
- Full resolution output — the effect is applied to the real image, not to the scaled-down preview you are dragging on.
- Works with touch — drag on a phone or tablet exactly as you would with a mouse.
- Undo and clear — step back one box at a time or start again without reloading the picture.
- Entirely offline — the photo never leaves your device, which is the whole point when the thing you are hiding is sensitive.
Frequently Asked Questions
Can the blurred area be recovered?
Not from the exported file — the pixels are overwritten before it is saved, so there is no hidden original. Be aware that a weak blur can still be read by eye, and that blur over a small, predictable set of values is theoretically reversible. Use pixelation at high strength or a black box for anything that really matters.
Why is this safer than drawing a black rectangle in a document?
Because a rectangle in a PDF, a slide or a design file is an object sitting on top of the content, and the content underneath is still in the file. Removing the rectangle reveals it. Exporting a flattened bitmap leaves nothing to remove.
Is the image uploaded anywhere?
No. It is drawn onto a canvas in your browser and the export is generated locally. Nothing is transmitted, which matters when the image is a bank statement or an identity document.
Which effect should I use?
Black box when the information must be unrecoverable and it is fine for the reader to see that something was removed. Pixelate at high strength for faces and number plates. Blur only for cosmetic softening — it is the weakest of the three.
Does it detect faces automatically?
No — you draw the boxes yourself. That is deliberate: automatic detection misses faces at an angle, partly hidden, or in the background, and a censoring tool that quietly misses one is worse than one that makes you look.
Does it remove the photo’s location data?
The export is written fresh from a canvas, which does not carry the original EXIF block forward, so GPS and camera fields are dropped. If you publish photos often, confirm it with Remove EXIF rather than relying on a side effect.
What formats can I use?
PNG, JPG and WebP go in. A JPG comes back as a JPG and everything else comes back as a PNG, so screenshots keep their sharp text and photos keep their file size sensible.