How to remove text from an image
Date stamps from disposable cameras. Captions burned into reposts. Watermarks on your own photos. Text overlays ruin otherwise good images. They're also one of the more solvable problems in photo editing, once you stop treating them as a pixel problem and start treating them as an object problem.
When you actually need to remove text from an image
There are a few common situations where text removal makes sense:
- Date stamps from film or disposable cameras. Your parents' vacation photos from 1998. Your own film photos with the date printed right across the corner. The photo itself is fine; the orange timestamp is the problem.
- Watermarks on photos you own the rights to. Stock photos you licensed, or your own photos that got reposted with someone else's branding on top. If you have the license or it's your original work, removing the overlay is legitimate cleanup.
- Text overlays from social media screenshots. Username bars, caption text, platform UI elements. Screenshots taken from Instagram or TikTok often have interface text baked into the image frame.
- Captions from reposted content. Someone shared your photo with their own caption text burned in, and you want the clean original.
One rule first: Only remove text you have the right to remove. Your own photos, images you've licensed, or content you own. Do not use text removal tools to strip copyright notices or watermarks from work that isn't yours.
How BOARD handles text removal
Most photo editors approach text removal as a masking problem: paint over the letters, run an inpainting algorithm, fix artifacts with the clone stamp. It works, but it puts the work on you. You're thinking about pixels.
BOARD takes a different approach. When you upload an image, it scans for objects: people, products, animals, and text elements. Date stamps, captions, and watermarks show up as selectable objects. You click the text you want removed, and BOARD fills the underlying area.
You never need to know which selection tool to use, what content-aware fill radius to set, or how to blend the edges manually. You just need to identify which text is the problem.
Four steps to remove text from an image
Use the original file if you have it. A JPEG exported at 80% quality works fine. Where things get harder: compressed screenshots, heavily filtered versions, or photos that have already been edited once. The cleaner the source file, the cleaner the result.
After upload, BOARD scans the image and identifies objects it can work with. Text elements (stamps, overlays, captions) show up as selectable regions. This takes a few seconds. You do not need to do anything during this step.
BOARD outlines detected objects when you hover over them. Click the date stamp or caption. You should see a highlight around the text confirming you have the right element selected. If the text spans multiple regions, select each one.
Hit remove. BOARD fills the area where the text was. Use the compare slider to check the result against the original. Most text removals look clean on the first pass. If there's a faint residual outline or mismatched texture, run a second pass on that region.
What works well vs what's harder
Usually works well
- Date stamps in a corner against sky or simple backgrounds
- Solid-color text overlays on photos with clear background areas
- Single-line captions at the bottom of a photo
- Watermarks printed over relatively uniform areas
- Text in high-resolution source files (more pixel data to work with)
Harder cases
- Text layered over complex textures: brick walls, foliage, patterned fabric
- Semitransparent watermarks that blend with the image below
- Very small or thin text where detection struggles to isolate the characters
- Large text blocks that cover significant areas needing reconstruction
- Text burned into a heavily compressed or low-resolution JPEG
Remove text from images without Photoshop
Photoshop can do this. The workflow involves the spot healing brush, content-aware fill, and sometimes manual clone stamping to blend edges. On a simple date stamp in a sky, an experienced Photoshop user might take three minutes. On complex backgrounds, it's longer.
BOARD compresses that to a click and a few seconds. It is not always as precise as a careful manual edit in Photoshop, but for most text removal jobs (stamps, social captions, basic watermarks) the automated result is fine and the time difference is significant.
If you need pixel-perfect control over the reconstruction, matching a very specific texture or blending into a complex pattern, Photoshop is still the right tool. For everything else, try BOARD first.
Frequently asked
Can BOARD remove date stamps from old photos?
Yes. Date stamps are one of the most common text removal jobs. They sit on top of the image, so the underlying area just needs to be reconstructed. BOARD handles this well on most backgrounds. Corner stamps against sky or grass almost always come out clean.
Does it work on watermarks?
Only remove watermarks you have the right to remove, meaning your own photos or images you own. BOARD can detect and remove text-style watermarks reasonably well, though semitransparent ones that blend into complex background areas are harder than solid text.
What about text that blends into the background?
Light text on a bright background, or dark text on a dark area, is harder for the detector to isolate. Try uploading a higher-resolution version of the image if detection misses the text. If the text truly can't be separated from the background, a manual approach in Photoshop may be the only option.
Does each text removal use one credit?
Yes. Each removal operation uses one credit, whether it's a person, an object, or a text element. Your first five edits are free, no account required.
Keep reading
Upload the photo with the text overlay.
Five edits free. No account, no watermarks on output. See the result in a few seconds.
Clean Up a Photo Free →