Home▸AI & Machine Learning▸2D Image Convolution: Edge Detection & Kernels (2D)

2D Image Convolution: Edge Detection & Kernels (2D)

2D image-convolution lab: a real pixel grid convolved live with Sobel, Gaussian, Laplacian and sharpen kernels — watch the kernel window slide across the source image while the output edge-map is computed pixel by pixel from the actual dot-product math.

AI & Machine Learning2DModerate60 FPS📱 Mobile-adapted⇄ 3D version

This 2D companion turns the discrete convolution behind edge detection into something you can watch happen: a generated pixel grid (circle, checkerboard, gradient, rings or noise) is scanned row by row by a real 3×3 or 5×5 kernel — Sobel X/Y, Sobel magnitude, Gaussian blur, Laplacian or sharpen — with the exact neighborhood values, kernel weights and resulting dot product shown live as the output edge-map is built pixel by pixel from the actual math, not a pre-rendered filter.

⚙ Under the hood

2D image-convolution lab with a real Sobel/Gaussian/Laplacian kernel sliding over a live pixel grid, computing and revealing the output edge-map pixel by pixel from the actual dot-product math.

image convolutionedge detectionsobel operatorgaussian blurkernel filtercomputer vision

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install