Object Remover
Erase unwanted people, objects, watermarks and blemishes with AI — paint over them and let AI fill the background. Free, unlimited.
Arraste sua imagem aqui ou clique para navegar
Supports PNG / JPG / JPEG / WEBP, up to 40MB
- Após pintar, uma imagem normalmente leva de 1 a 5 minutos
- Os resultados ficam disponíveis por 24 horas e depois são excluídos automaticamente
- Sua tarefa é salva automaticamente: você pode fechar a página e voltar em 24 horas (apenas a última tarefa é mantida; evite modo anônimo e não limpe o cache do navegador)
Quando o processamento terminar, o resultado será enviado para este e-mail (apenas uma vez)
Paint over areas to remove
Paint over the people, objects, watermarks or blemishes to remove — AI will fill the background. Cover the area fully for best results.
O resultado aparecerá aqui após pintar e enviar
The AI behind object removal from photos
Removing unwanted objects from a photo is powered by a diffusion-based generative model. First, an encoder compresses the whole image into a highly abstract latent space, where the model learns the structure of the scene, occlusions and semantic relationships — for example, recognizing that a tree is standing in front of a building. When you paint over the object you want to remove, the model repeatedly runs a noise-add-and-denoise diffusion process, rebuilding the pixels of the painted region layer by layer so the newly generated background matches the texture direction, lighting and perspective of its surroundings. No manual masking or cloning is needed: pixels outside the painted area are preserved exactly, and the final output is a clean photo with no visible traces of editing.
When to use the object remover
Use this free online object remover for many everyday cases: erase bystanders and tourists from travel photos so the scene looks clean and intentional; remove power lines, lampposts or trash bins that break the composition and let your subject stand out; delete date stamps, timestamps and brand logos so the image is ready to share or reuse; clean up blemishes, spots and marks on portraits for a more polished look; and repair scratches, stains and creases on old scanned photos. Everything runs online — no software to install, just paint and remove, with results available to download within 24 hours. No account or payment is required, so you can remove a person from a photo or clean up a watermark as often as you like, for free.
FAQ
Is this tool free?
Yes, completely free with no sign-up and no usage limits.
Which image formats are supported?
PNG, JPG, JPEG and WEBP, up to 40MB per file.
How do I use it?
Upload an image, paint over the people, objects, watermarks or blemishes you want to remove, then click start — AI fills the background of the painted area automatically.
What can it remove?
Bystanders and tourists, power lines, clutter, date stamps, watermarks, text, skin blemishes, and scratches or stains on old photos — paint over anything you do not want.
What happens to the background after removal?
AI understands the structure of the scene around the painted area and generates natural textures, lighting and perspective so the fill blends seamlessly with the surroundings.
What is the output format?
A processed JPG image (high-quality compression, smaller file for faster loading), ready to download.
How long does processing take?
Usually 30 seconds to 3 minutes per image after painting (the AI rebuilds the removed region layer by layer).
Can I paint multiple areas at once?
Yes. All painted areas are combined into a single mask and processed by the AI in one pass: the model sees every region to remove together with the context of the whole scene, so it fills them in a coordinated, natural way. Pixels outside the painted areas keep the original sharpness — no repeated processing or visible seams.
Is my file stored?
No. Images are automatically deleted within 24 hours after processing.
What if I paint the wrong area?
The painting editor provides an eraser and undo, so you can adjust your mask anytime — or clear it and start over.
What technology does it use?
A deep-learning diffusion generative model: it first understands the structure and semantics of the whole image in a highly compressed latent space, then denoises step by step to rebuild the pixels of the painted region, keeping texture, lighting and perspective consistent with the surroundings.