How do I remove a watermark from an image?

Use Imgour's object remover. Upload your image, brush over the watermark, and AI removes it instantly.

How do I remove the Gemini logo from an image?

Go to www.imgour.com/remove-gemini-logo, upload your image, and our AI will erase the Google Gemini watermark or logo instantly.

What AI image tools does imgour offer?

imgour offers object and background removal, watermark and logo removal, AI headshot and product photo generation, pet portraits, talking photos, AI song covers, and adding sound to video.

How do I remove the background from an image?

Use Imgour's Background Remover. Upload your photo and AI automatically removes the background, creating a transparent PNG.

How do I generate an AI headshot?

Use Imgour's AI Headshot Generator. Upload a few selfies and AI turns them into professional studio-quality headshots.

For the complete documentation index, see llms.txt or the full corpus at llms-full.txt.

AI Photo Cleanup

Remove the clutter, wires, litter, and distractions that ruin an otherwise good picture. Brush it, and AI rebuilds what was behind it.

Free and unlimited. Your subject is never redrawn.

Upload picture

or drop photos here

How to Remove Something from a Picture

Upload the picture

Any common format works. Images above 2000px on the longest side are scaled to 2000px before editing.

Brush what you want gone

Cover the object fully and go a little way past its edges. A rough, generous stroke beats a careful outline — the fill is inferred from the pixels bordering your brush, so it needs a margin of clean background to read from.

Work in passes on messy shots

If the picture has several distractions, do not brush them all at once. Remove the easy ones first. Each completed pass gives the next one more clean surface to sample, which compounds into a noticeably better final result.

Download

No watermark, no credit card required.

What Usually Needs Cleaning Up

Power lines and cables

Wires across a sky are the textbook case. They are thin and sit on a smooth gradient, so they disappear almost perfectly.

Litter, bins, and street furniture

Rubbish on a beach, a wheelie bin in a garden shot, bollards and signposts in an architectural photo.

Clutter in interior photos

Cables under a desk, a drying rack in a room shot, items on a worktop. The surrounding floor or surface is usually uniform enough to rebuild convincingly.

Blemishes and marks

Spots on skin, scuffs on a wall, stains on fabric, dust on a lens. Small and surrounded by similar texture — ideal conditions.

Damage on scanned prints

Creases, scratches, tape residue, and specks on old photographs.

Reflections and glare spots

A photographer reflected in a window or a chrome surface, or a small hotspot on a glossy product.

Cleanup Is Not the Same as AI Image Generation

If you ask a general-purpose AI assistant to “remove the bin from this photo”, it will typically generate a brand new image based on yours. The bin disappears, but so does the exactness of everything else: faces drift, writing on signs becomes unreadable, colours shift, and the output is often returned at a smaller size than the original.

Cleanup works the other way round. The image is treated as fixed, a mask marks the one region to rebuild, and only that region is regenerated. Everything outside the mask is your original data, unchanged. That is why cleanup is the right tool whenever the photo has to remain recognisably the same photo — a property listing, a product shot, a document scan, a family picture.

It is also the reason cleanup handles high-detail images well while generative editing struggles with them: nothing is being re-imagined except the small area you chose.

Getting a Cleaner Result

Brush wider than the object. Include a margin of background. Tight tracing is the most common cause of a visible halo left behind.
Take the shadow too. Removing an object does not remove the shadow it casts or its reflection. Both need their own strokes.
Split large removals. One enormous brushed area gives the AI very little to infer from. Several smaller passes give it progressively more.
Start with the simple background. Where an object crosses both plain sky and detailed foliage, remove the sky portion first.
Accept the hard cases. Anything sitting in front of unique, unrepeated detail — readable text, a face, a specific pattern that appears nowhere else — has no reference to rebuild from and will need manual retouching.

Frequently Asked Questions

How do I remove something from a picture?

Upload the picture, brush over the thing you want gone, and the AI rebuilds the area behind it from the surrounding pixels. It works for objects, people, text, marks, and clutter — the method is the same regardless of what you are erasing.

What counts as photo cleanup?

Anything you remove rather than add: power lines across a skyline, litter on a beach, a bin in a garden shot, dust on a scan, a stain on a wall, or a stray hand at the edge of frame. Cleanup means taking distractions out and leaving the subject untouched.

Is AI image cleanup the same as using ChatGPT to edit a photo?

No, and the difference matters. A general AI image generator redraws the whole picture, so faces shift, text turns to nonsense, and the result is a new image that resembles yours. Cleanup regenerates only the area you brushed and passes every other pixel through unchanged.

Can I clean up several things in one photo?

Yes. You can brush multiple separate areas before removing, or work in passes. Passes usually give better results on cluttered photos, because each completed pass restores clean surface that the next one can sample from.

Does cleanup work on old or scanned photographs?

Yes. Scratches, creases, dust specks, tape marks, and foxing on scanned prints are good candidates — they sit on top of a fairly uniform paper or emulsion surface, which reconstructs predictably.

Will the cleaned photo look edited?

On plain and repeating surfaces, generally not. On unique, non-repeating detail directly behind the removed object the AI has to invent content, and that is where an edit becomes noticeable. Removing in smaller pieces helps considerably.

What resolution do I get back?

Images are processed at up to 2000px on the longest side. Anything larger is scaled down to that before editing, so a very large original comes back smaller. There is no watermark on the result.