Images
Align multiple photos and remove passing people, cars and other moving objects on your device.
Photos are never uploaded. All image processing runs on this device.
Photos are automatically aligned. A machine learning model detects the positions and outlines of people, animals and vehicles. Comparing the photos identifies moving objects, whose regions are excluded and replaced with background actually captured in other photos. No generative AI is used to create or invent image content. Background hidden in every photo cannot be recovered, so those areas retain the original image. Start with 3–5 photos. If objects remain, adding photos that show the background can improve the result. More photos take longer to process.
Use at least 3 photos taken from the same position and with the same framing. Start with 3–5 photos. Keep the camera position, zoom and orientation steady, and wait for people or vehicles to move between shots. Lock exposure and focus if possible.
Adding photos that reveal background previously hidden by people or vehicles can improve removal. Adding similar photos with objects in the same positions does not guarantee improvement. You can select up to 20 photos, and more photos take longer to process.
The tool detects and removes people, animals and vehicles moving between photos. Stationary people, parked vehicles and objects occupying the same position in most photos may remain. Background hidden in every photo cannot be recovered. Flags, leaves, water, reflections, changing shadows and differences in camera position can also cause artifacts. This tool is in beta. Check the finished photo, Before / After comparison and alignment overlay before saving.
A machine learning model detects the positions and outlines of people, animals and vehicles. No generative AI creates or invents image content. Regions identified as moving are excluded and replaced with background actually captured in other photos.
Photos are never uploaded. Alignment, detection and compositing all run on your device. An internet connection is needed initially to download resources, including about 25.5MB of detection data (model and runtime combined).
Start with Fast (up to 1920 px on the longest edge) to check the result, then try High quality (up to 3840 px) for saving. Switching to High quality and running again with the same photos reuses the previous analysis to reduce processing time. Photos are never enlarged beyond their original resolution. Adding or removing photos, resetting or reloading the page requires a new analysis. Processing time depends on your device and photos.
Supported EXIF fields from the reference photo, including capture dates and GPS location, are retained in the saved JPEG when available and readable. Orientation and dimensions are updated for the output, embedded thumbnails are removed, and Software is set to LOUMO TOOLS. Your original photo files are not modified.
Open this page online and wait for the offline-ready message. For the first use, also select at least 3 photos, start processing and accept the detection-data download. Once composition has completed, you can use this page to process photos offline. Saved detection data remains after you close the page. Private browsing, clearing browser data or automatic storage eviction may remove it. If that happens, open the page online again and download any missing resources.