Browser-local electronic stamp workspace
Extract a Transparent Stamp from an Image
Drop in an authorized stamp image, keep the required color information, and remove its background locally in your browser.
- Drop or choose an image
- Adjust retained range and cleanup
- Transparent background and edge enhancement
- Return the result to the editor
Try the downloadable example
This computer-drawn ring contains no real company seal or document. A red bar and a small red dot were added to demonstrate cropping and manual cleanup.
Choose “Open the stamp extractor” to enter the tool directly. Download the example below and drop it into the extractor.

Step 1: Crop around the subject

- After uploading, drag on the source preview around the ring with a little padding, excluding the red bar at the top left.
- Release to re-extract automatically. Use the full-image button to reset a mistaken selection.
Step 2: Adjust settings, then erase noise

- Make small changes to retained range and cleanup, keeping the ring or lettering complete.
- Choose Erase and use a small brush on the isolated dot. Undo a stroke or use Restore to recover accidentally erased pixels.
- Finish crop and extraction settings before retouching; re-extraction resets the current retouch work.
Step 3: Return to the canvas and export PNG
JPEG does not preserve transparency. Current SVG output embeds a PNG bitmap and does not support unlimited lossless scaling.
- Keep Transparent Background selected and choose Replace Canvas. This replaces the current canvas; editor Undo can restore it.
- Open Download, choose PNG, and leave the white-background option off.
- For a specific print size, enable physical-size export, enter stamp width and DPI, and print at 100% original size. The image includes extra edge padding.
Troubleshooting the result
Missing strokes or faint edges: reduce cleanup, adjust the retained range, and check the source quality. Restore cannot bring back pixels already discarded during extraction.
Remaining red text, lines or dots: tighten the crop first, then use a small brush on isolated noise.
Heavy occlusion, strong shadows or no distinct red subject: try a clearer source. The tool does not infer or reconstruct invisible content.
Only process images you own or are authorized to use. This workflow does not verify seals, documents or permissions.