There are three common โ™ป๏ธ Convolution kernels in ๐Ÿ‘๏ธ Computer Vision, all based on the ๐Ÿ‘‘ Gaussian distribution

  1. Gaussian: gaussian fixed at the center, blurs image.
  2. Derivative: derivative of gaussian, detects directional differences in image.
  3. Laplacian: second derivative of gaussian, detects details in image (like sharpening).