Seven useful algorithms
Choose the edge detector that fits the job, from Canny for photographs to region contours for logos and flat graphics.
Turn images into clean outlines, high-contrast edge maps and scalable vector contours. Tune the detection, compare results and export—all in your browser.
YOUR CREATIVE WORKSPACE
Start with any image, then dial in your perfect outline.
01 · SOURCE
02 · DETECTION
03 · FINISH
OUTPUT QUALITY
LIVE PREVIEW
Upload an image or try the sample to preview its edges instantly.
EXPORT STUDIO
Raster PNG or editable vector outlines. No watermark.
Start with Canny for photographs. Try Contour mode for logos or solid shapes, and increase the threshold to remove distracting texture.
MAKE MORE WITH PIXELOKAY
Practical tools for makers, designers and developers—without uploading your work.
Choose the edge detector that fits the job, from Canny for photographs to region contours for logos and flat graphics.
SVG export follows the pixel-boundary outlines to create fillable, scalable geometry. Dense images may need lower resolution to keep file sizes practical.
Images are decoded and processed entirely in your browser using an isolated Web Worker. No files are sent to an external server.
Canny provides thin photographic edges with dual thresholds. Sobel, Scharr and Prewitt highlight gradients, Roberts detects rapid changes in small neighborhoods, and Laplacian emphasizes high-contrast changes. Contour traces the selected dark or light regions.
Yes. PixelOkay turns the detected edge-pixel boundaries into SVG path geometry; it does not embed a bitmap. Very detailed outlines create more path data, so higher thresholds or a smaller working resolution may help.
No. All image loading, processing and export take place on your device. The app does not transmit image content. Your browser handles any files you download.
The exported geometry is based on pixel boundaries, not smooth centerline tracing or machine-tool calibration. Check the paths, scale and tolerances in specialist software before manufacturing.
Create local raster edge maps and geometric contour outputs by tuning thresholds, filtering components and reviewing the resulting regions.
Use a high-contrast object, adjust the edge threshold, remove small components and compare the transparent raster with the SVG paths before exporting both.
This uses image-space edge and region algorithms, not semantic object recognition or manufacturing-grade measurement. Thresholds, component filters and dilation change the geometry. Raster and SVG agreement needs independent review. Image processing is local and no remote model is used; actual output fidelity has not been independently accepted.
No. It detects image edges and regions using adjustable image-processing operations.
No. They are image-derived geometry, not calibrated or manufacturing-certified measurements.