Background Remover

This tool works by colour, not by recognising objects. It looks at the colours along the edges of your photo, treats anything that matches as background, and removes it. That makes it well suited to product photos on a plain backdrop, logos and graphics on a solid colour, and studio-style portraits against a wall, and less suited to a person standing in a busy street or a subject that shares its colour with the background. How it works. When you add a photo, the tool samples the colours along its four edges and groups them into a small number of clusters, since a real backdrop is rarely a single flat colour: studio lighting shades it, and JPEG compression adds noise. It then removes the region connected to the edges that matches those colours, working inward so an object of a similar colour inside the photo is not touched by mistake, unless you turn on “Also remove enclosed areas”. Around the boundary it works out how much of each edge pixel is background and how much is subject, and recovers the true colour of the subject portion. That is what keeps a plain background from leaving a faint white or grey fringe around hair, fur, or a curved edge. Adjusting the result. Sensitivity controls how close a colour has to be to the detected background to count as background: raise it if patches of a shaded or gradient background are left behind, lower it if parts of the subject start to disappear. Edge softness blends the cut smoothly, which matters for anything with soft or slightly blurry edges. Shrink or grow the edge lets you cut a pixel or two into the subject to remove a stubborn fringe, or grow it back out. “Remove light shadows” treats a soft shadow cast by the subject onto the background as background too, which suits product photos with a drop shadow beneath them. When the automatic result needs a hand. Use “Pick background” to click a colour you want removed, such as a second backdrop colour a corner picker missed, or the inside of a shape that the edge detection couldn't reach. Use the erase and restore brushes to paint parts of the result in or out by hand, with undo and redo. These tools handle situations that colour-based detection cannot, like an off-colour object in the scene or a background that is not connected to the edge of the photo. New backgrounds. Once the subject is cut out, place it on a solid colour, a blurred version of the original photo (a quick way to keep some context without the distraction), or another image of your choosing, resized to fill the frame. Or keep it transparent as a PNG or WebP file. Batch processing. Add up to 50 images, and the detection settings and new background apply to all of them, while brush edits apply to the image you're working on. One image downloads directly; several download together as a ZIP. Output. Save as PNG (always supports transparency), WebP (smaller files, transparency where your browser supports it), or JPG (no transparency, so a white background is used automatically). You can crop tightly to the subject with adjustable padding, and limit the longest side of the output for smaller files. What this tool is not. It does not use an AI model to recognise people, animals, or objects, so it cannot separate a subject from a background that is not a fairly plain colour or gradient, such as a busy room or outdoor scene. For those photos, expect to use the pick and brush tools, or a result that needs more manual cleanup. Privacy. Your photos are processed by your browser and are not uploaded to EaseBowl's servers. EaseBowl records basic usage statistics, such as how often a tool is used, but not your files or their contents.