Photo Privacy Blur & Redaction — Manual Regions to PNG

Before posting a screenshot or photograph, you may need to hide a phone number, address, badge, face or other identifying detail. This Photo Privacy Blur and Redaction tool lets you select rectangular areas and create a flattened PNG with those areas covered. It offers solid cover, pixelation and blur, with solid cover selected by default because it replaces the visible pixels inside the chosen rectangle.

Selection is manual. The tool does not find faces, read text, locate personal details or decide what needs hiding. The person using it must choose every relevant area and inspect the downloaded result. Blur and pixelation are visual effects that can leave information recognizable; a softened region should never be taken as proof that a sensitive detail cannot be inferred.

How to cover an area

Choose a JPG, PNG or WebP photo. Select a method, then enter the rectangle's left edge, top edge, width and height as percentages of the image. Left and top start at zero. Width and height must be positive whole numbers, and the rectangle must fit within the picture. Create a preview to see the position and coverage, then adjust the fields if needed.

For example, left 25, top 20, width 50 and height 35 covers the central half of the picture's width, starting one fifth of the way down. On a 1000 × 800 output, that means an area around x=250, y=160, width=500 and height=280 pixels. Percentages are calculated from the whole image, not the visible browser box. This means the same relative area remains selected even when the preview is scaled to fit a phone.

Working with several rectangles

For multiple details, enter the first rectangle and press Add this region. Change the fields for the next area and add that region too. You can store up to 20 rectangles. Each stores its own method at the time it is added, so changing the method later does not change earlier entries. Create a fresh preview after building the list.

When the added list contains any rectangles, the tool applies that list and ignores the current fields until you add them. This avoids applying an unconfirmed edit alongside saved selections. If the list is empty, the current fields define one rectangle. Clear added regions empties the list; the fields remain and can then be used for a new single-area preview. Changing settings removes old output links so you must regenerate before downloading.

Choosing solid cover, pixelation or blur

Solid cover paints an opaque dark rectangle over the selected area. The downloaded PNG contains those painted pixels instead of the original visible pixels inside that rectangle. There are no editable image layers in the export. However, an uncovered edge, reflection, repeated detail elsewhere in the photo or the surrounding context can still disclose information. Leave enough room around the entire sensitive item.

Pixelation reduces the region to coarse blocks and scales those blocks back up. It can help create an anonymous-looking illustration, but recognizable shapes or text may remain. Blur uses a Canvas blur filter at a fixed strength. Large text, distinctive faces or insufficiently covered areas may still be recognizable. Browser support for the filter varies; use solid cover if your browser cannot apply blur. Neither effect is a guarantee of anonymization.

Practical examples before sharing

For a delivery label photo, cover the complete name, address, phone number, barcode and tracking details that you do not want public. A rectangle over only one line may leave the same information visible elsewhere. Look at the whole image and consider whether a cropped copy would remove unrelated material more clearly. Keep the original private if it contains details you still need for your own records.

For an event picture, inspect name badges, screens, vehicle plates and background signs in addition to the main subject. For a screenshot, account names can appear in headers, notifications or quoted messages. This tool operates on still images, not a whole video or every page of a document. If you have multiple images, process and check each separately rather than assuming one set of percentages will cover every source correctly.

Output dimensions, file limits and metadata

Inputs must be genuine JPG, PNG or WebP files, at most 25 MB and at most 24 megapixels after decoding. The output preserves the aspect ratio, limiting the longest edge to 2000 pixels without enlarging smaller images. A 4000 × 3000 photo becomes 2000 × 1500. Rectangle coordinates are applied to that rendered output so their percentage positions remain consistent after resizing.

The download is a flattened PNG named redacted-photo.png. Canvas creates a new image from rendered pixels; the original file is not modified. This tool does not provide an EXIF inspection report or a forensic metadata audit. If metadata verification matters to your workflow, check the new file with an appropriate metadata tool as a separate step. Transparent areas outside the selected rectangles can remain transparent in PNG output.

Open the downloaded file in a separate viewer and zoom into each covered area. Check rectangle edges, spelling-shaped remnants, similar information elsewhere and details that become visible against different backgrounds. Do this with the actual saved PNG, not only the scaled preview. The page does not publish your picture or send it to a messaging app; sharing is a separate action you choose after checking the file.

Privacy, browsers and common problems

Image decoding, rectangle rendering and PNG creation happen locally in the browser. The tool has no request that uploads your selected photo and no server-side editing history. The website may separately load its ordinary advertising, analytics or page resources. Refreshing clears the current session, so save before leaving. Use a recent browser with JavaScript and the required image decoder available.

An “outside the photo” message means left plus width, or top plus height, exceeds 100. Reduce the rectangle or move its start. A failed image may be damaged, unsupported or larger than the decoded-pixel limit. Try a smaller supported copy. If download behavior is restricted inside a social app, open the page in a normal browser. The percentage fields work with touch and keyboard entry; this version does not offer drag selection or automatic detection.

Frequently asked questions

Is blur enough to hide sensitive information?

It may not be. Blur and pixelation can preserve recognizable clues. For the strongest visible-pixel removal offered here, choose solid cover and cover the complete detail with suitable margins, then inspect the saved file.

Can I cover several faces or numbers?

Yes. Add up to 20 rectangles, each with its selected method, and create a new preview. When the list is not empty, only its stored rectangles are applied; add the current fields if you want them included too.

What file is downloaded?

You receive a flattened PNG with a longest edge of at most 2000 pixels. Smaller sources are not enlarged. The original is unchanged, and the PNG has no editable layer containing the covered source region.

Does the tool upload or automatically detect my details?

No. Its image processing is local and selection is manual. Normal page resources may still use the network. The tool does not detect personal information or certify that the entire image is safe to share.

Can I use it on mobile or on a video?

The percentage controls adapt to mobile browsers, subject to file and download support. This version accepts still JPG, PNG and WebP images. It does not redact video frames over time or process a multi-page PDF.

Why does my rectangle look misplaced?

Coordinates describe percentages of the whole image, not screen pixels. Check left, top, width and height, then regenerate. If you have stored rectangles, changing current fields alone will not alter them; clear the list or add the new region.