Pixelate & Blur Image

I built this after nearly posting a screenshot that had a full account number sitting in the corner of it. I caught it on the second look, which is one look more than most people get. The problem is that censoring an image usually means opening a proper editor, and nobody opens a proper editor to hide one line of text before sending a screenshot. So this does exactly that one job: drag a box over anything that should not be visible, pick pixelation or blur, set the strength, and download a censored copy.

You can add as many boxes as you need, over faces, license plates, street signs, addresses, email headers, card numbers, or a colleague's name in a chat log. There is an undo if a box lands in the wrong place. The image is drawn and re-drawn on a canvas inside your browser, so the original never leaves your device, which is a strange thing to have to say about a privacy tool and yet most of the competition uploads first and censors second. If the photo is also carrying location data you would rather not share, run it through the EXIF remover afterwards.

How to use

1. Drop your photo or screenshot onto the box, or press the button and choose a file.
2. Choose pixelation for the blocky look, or blur for a softer smear.
3. Set the strength. Stronger means larger blocks or a heavier blur and less chance of anything being readable.
4. Drag a box over the first thing you want hidden. The effect applies inside that rectangle only.
5. Keep adding boxes until everything sensitive is covered. There is no limit on how many you draw.
6. Use undo if a box is in the wrong place or the strength was not enough.
7. Zoom in and check the result at full size, then download the censored copy.

Why use our pixelate & blur image?

The important thing here is that the censoring is real. When you download, the pixels underneath your boxes are genuinely gone from that copy, replaced by the blocks or the blur you see. This is not the same as drawing a black rectangle over text in a PDF viewer or a slide, which leaves the original content sitting under a shape that can sometimes be moved, deleted, or read straight out of the file. That mistake has embarrassed law firms, government departments and newspapers. Here the exported image is flattened, so what is hidden is hidden in the file itself.

Pixelation and blur are both offered because they fail differently. Pixelation replaces a region with large flat blocks, which is obvious to a reader. Blur smears detail and looks tidier, but a weak blur leaves a surprising amount of structure behind, which is why the strength control exists. The blocky effect comes from scaling a region down and back up with image smoothing turned off, which gives real, hard edged blocks rather than a cosmetic filter.

Now the honest caution, because a privacy tool that oversells itself is worse than no tool. Use a strong setting for anything that matters. Light pixelation over a short string of known characters, say a six digit code, leaves less doubt than you would think, because a determined person knows the shape of the alphabet they are looking for. Go heavy on numbers, names and plates. If the thing is truly sensitive and you can afford to lose that part of the picture, cropping it out with the crop tool is the strongest option of all, because the data is simply not in the file.

The part I got wrong first was a low default strength, so the preview looked gentle and pretty. It looked good and it was quietly dangerous, because a soft default is exactly what someone in a hurry accepts. Now the strength control is front and center, and the guidance is to overdo it. Nobody has ever regretted a redaction that was too thorough.

What I refuse to add is automatic face detection. It sounds helpful and it is a trap: detection misses the person in profile, the reflection in the window and the face at the back of the frame, and it teaches you to trust a machine with a job that needs your eyes. Dragging boxes takes seconds longer and you know exactly what you covered. There is also no cloud saving and no account, so the uncensored original never exists anywhere but your device.

Who is this tool for?

Screenshots are the daily case. Support teams share a customer ticket without the customer's email address. Developers post a console error without the API endpoint. Anyone sharing a bank app screen to prove a payment covers the balance first. Sellers hide their address on a photographed shipping label before posting proof of dispatch.

Photographs come next. Parents blur other people's children before putting a school event photo in a group. Property listings hide a neighbor's plate. Journalists protect the identity of a source or a bystander. Anyone posting a picture of their street wants the house number gone before it goes public.

Documents need the same treatment and the route is slightly longer. Turn the page into a picture with PDF to JPG, censor it here, and you have a redacted page whose hidden content is genuinely destroyed rather than merely covered up, which is the failure mode people keep discovering the hard way.

Businesses use it for case studies and portfolios: showing a real dashboard, invoice or client email while hiding the names and figures that made it real. That is the trick of publishing convincing evidence without leaking anything.

Frequently asked questions