Blog
Log in

Remove People and Passersby from Photos (AI LaMa)

Remove a tourist, crowd, or photobomber from your shot and keep the landmark and landscape. A specialized workflow on the LaMa neural network with hints tailored to people.

Remove people from photos
Scenario

Wide brush: cover the silhouette from head to toe, including the shadow on the ground.

Drag a photo with bystanders here

Cover the figure entirely with a wider brush and click "Remove person". Better one person at a time.

Ready - open a photo with people

Instructions

  • Upload a photo of a landmark, street, or landscape with unwanted people.
  • Choose a scenario: one person, a crowd, or a photobomber at the edge.
  • Use a wide brush to cover the figure entirely (and the shadow).
  • Click Remove Person. For a crowd, repeat in segments.
  • Save the PNG and keep the focus on the architecture or scenery.

Повний опис

Specifically about people in the frame, not just an object remover

A general object remover can take out anything. This service is built around one request: keep a landmark or landscape without tourists, queues, and photobombers. Brush presets are wider, hints cover silhouette, shadow, and working one person at a time, and the text is about travel and architecture photos.

Typical cases

  • A cathedral, square, or museum without random selfies in the foreground.
  • A landscape or embankment without groups in bright clothing.
  • Wedding or couple shots: remove a passerby from the background (photobomber).
  • Commercial location shoot: a clean frame for a hotel or guide website.

Strategy for crowds

Don't cover half the frame with one mask: LaMa restores the background better when it sees enough context around it. Remove people one at a time or in small groups. Include the shadow on the pavement in the mask, otherwise a ghost will remain.

If a person blocks a complex grid or ornament, do several passes and check the ornament at zoom.

Frequently asked questions

How is it different from the general object remover?
The same LaMa engine, but different brush presets, statuses, and texts tailored to the case of removing people / tourists / photobombers. It is more convenient for travel and architecture photos when the task is specifically about people in the frame, not an arbitrary object.
Why is it better to remove people one at a time?
This way the background (cobblestones, facade, grass) is restored more reliably: the model sees more context around the mask. A large mask covering half a crowd often produces a smeared seam. Work on 1 or 2 figures per pass.
A person's shadow is left on the ground
Add the shadow to the mask and repeat the pass with the One person scenario. An unpainted shadow looks like a ghost and gives away the retouch. If the result is not perfect on the first try, refine the mask while zoomed in and repeat with a short pass.
The person is partially hidden behind a column or a tree
Paint only the visible part of the figure; do not include the architecture or the tree crown you want to keep. Check the edge of the column while zoomed in after the pass. If the result is not perfect on the first try, refine the mask while zoomed in and repeat with a short pass.
Which image formats are supported?
PNG, JPEG, WebP, and other formats that the browser can open are accepted as input (usually BMP and GIF as well). The result is always saved as PNG to avoid repeated JPEG compression after retouching and to preserve the edges after inpainting without extra compression artifacts.
Where is my image processed and is it stored?
You draw the mask locally in the browser. Inpainting (the LaMa neural network or classic OpenCV) runs on the filltools.com server only to process your request. The file is not published in the gallery and is not intended to be shown to other users after the session ends; do not upload unnecessary personal data without need.
How is LaMa different from blurring or an eraser?
Blurring and pixelation only hide an area with a gray or pixelated patch. LaMa (Large Mask Inpainting) analyzes neighboring pixels and synthesizes a plausible background, such as sky, wall, fabric, or paper, so that the edge looks more natural. For small edits, the fast OpenCV Telea or Navier-Stokes option in the method list is sometimes enough.
Can I undo a mistaken step?
Yes. The Undo and Redo buttons keep a history of the mask and results (roughly up to 20 steps in the current browser session). Reset mask clears only the red markings on the canvas; Reset fully returns the tool to the state with no uploaded image.

Reviews (0)

Similar tools

Other tools from this category you may need next.

{ }

AI Image Upscaler (Real-ESRGAN)

Upscale photos 2× or 4× without blurry pixels. Quality mode uses server-side Real-ESRGAN; Fast mode…

1
{ }

Remove Image Background Online (AI BiRefNet)

Remove backgrounds from photos with the BiRefNet neural network or the classic JS mode. Mask editor…

1
{ }

Remove License Plates from Cars Online (AI LaMa)

Paint over the license plate and the LaMa neural network will draw the bumper or grille in its plac…

0

Blog

Articles about tools, development, and practical tips.

All articles