The short answer
Why use Edge Detection?
Outlines are useful for tracing a photo into a drawing, checking which parts are in focus, making a neon-style graphic, or preparing line art. A Laplacian filter does it in one pass instead of tracing by hand.
Edge strength comes from Amount, 0 to 100, which opens at 50. The filter is the 3 × 3 Laplacian: each pixel counts eight times, minus each of its eight neighbors. Where all nine match, the result is zero, which is black. Amount multiplies the weights, so 100 doubles the edge brightness and 0 turns the whole image black.
The Laplacian reacts to every small change, including JPEG blocks, sensor noise and fine fabric, which show up as sparkle across the picture. Smooth the photo slightly with Blur at a low Amount first for cleaner lines. Transparent areas stay transparent, so a cutout keeps its shape.
Pencil Sketch gives dark lines on white, closer to a drawing, and prints far better. Emboss shows edges as a raised gray relief. Run Invert after Edge Detection if you want dark, colored lines on a white background instead of light lines on black. Both steps keep the full resolution of your photo.
Where Edge Detection fits
Vinyl cutting
Get the outline of a logo to import into software for a vinyl or paper cutter.
Focus checking
See which parts of a product shot are sharp by comparing how bright their edges are.
Neon artwork
Turn a city skyline photo into glowing colored lines on black for a poster.
Start to finish
How to use Edge Detection
- 01Tap Choose image and pick a sharp photo or drawing.
- 02Raise Amount above 50 if the outlines look faint, or lower it if fine noise sparkles everywhere.
- 03Tap Apply changes, check that the main shapes are traced, then tap Download result.
What to expect from Edge Detection
| Check | What it means here |
|---|---|
| Supported input | JPEG, PNG, WebP, GIF, BMP, SVG and browser-supported AVIF images |
| Main controls | Amount slider, plus a color picker for effects that need one |
| Output | A PNG image with the effect applied |
| Working limit | 50 MB and 50 megapixels per image |
Edge Detection tips
- Blur the photo lightly at Amount 5 to 10 first to remove sparkle from JPEG noise.
- Run Grayscale on the result for white-only lines.
- Run Invert after for dark lines on white that are easier to trace on paper.
- Amount 0 gives a fully black image, so never leave it there.
Edge Detection FAQ
Something missing? Ask us.
Why are the edges colored?
Why is the result mostly black?
Does Edge Detection upload my file?
Is Edge Detection free to use?
Why did Edge Detection fail or look wrong?
Related tools
More in Filters & Effects.