All tools
Transform & Edit

Background Remover

Remove solid-color backgrounds using smart flood-fill detection

Powered by @imgly/background-removal — a full ONNX neural network that runs entirely in your browser. Works on any image type (people, products, hair, complex objects). No data leaves your device. First run downloads the AI model (~50 MB), subsequent runs use the browser cache.

Drop an image here or click to browse

Accepted: JPG, PNG, WebP, GIF, BMP, TIFF, AVIF · Max 1 file · 100 MB per file

About this tool

Remove a solid-colour background from an image to produce a transparent PNG. Best for product shots on a uniform white or coloured backdrop, logos against a single colour, or simple isolated subjects. Uses a flood-fill detection approach — fast and runs entirely in your browser, no AI service required.

When to use it

  • Producing transparent product cutouts for an e-commerce site
  • Isolating a logo against its solid-colour backdrop for use on different surfaces
  • Cleaning up a screenshot's background for use in design layouts
  • Producing transparent thumbnails from a uniform background
  • Preparing simple isolated illustrations or icons for layered designs

What to expect

Flood-fill works well on solid or near-solid backgrounds. Complex scenes (gradients, busy patterns, fine details like hair) need an AI-based background remover for best results. The algorithm starts from the corners and grows outward; tweak tolerance for slight colour variations in the background.

Frequently asked questions

Will this work on photos with complex backgrounds?

Not reliably. This tool uses flood-fill detection, which excels at solid or near-solid backgrounds. For photos with detailed scenes or fine subject edges (hair, fur), a dedicated AI background remover produces cleaner results.

What format does the output come in?

PNG with a true transparent alpha channel — usable on any coloured background. The transparency is preserved in any tool that supports PNG (web browsers, design software, document tools).

Why are some background pixels not removed?

Slight colour variations in the source can leave isolated patches the algorithm doesn't include. Increase the tolerance slider to be more aggressive, or pre-process the image to make the background more uniform.

Related image tools