Scharr Filter
A refinement of the Sobel kernel developed by Hanno Scharr around 2000 to estimate gradient direction more accurately. It gives cleaner, more rotation-consistent edge maps on small 3x3 neighborhoods, useful for orientation-sensitive analysis.
Spot it
- 01Crisp bright edge lines
- 02Even strength on diagonals
- 03Dark quiet background
Why it happens
Uses a 3x3 derivative kernel with 3, 10, 3 weights across the smoothing direction, chosen to minimize angular error so edges at every orientation respond equally.
Recreate it in Pixio
Image edit prompt“Convert this image into a Scharr gradient map: a dark, nearly black background with sharp bright white edge lines that respond equally on diagonal, vertical and horizontal contours. Fine details such as hair and fabric texture appear as crisp thin strokes. Grayscale only, high precision look. Keep the original subject and composition recognizable.”