Kernel

class Kernel(K: ndarray, name: str | None = None)

Convolution kernel object

Parameters:

• K (ndarray(N,M)) – kernel weighting matrix

• name (str, optional) – name of the kernel, defaults to None

Raises:

ValueError – K is not a 2D ndarray

Kernel objects are used to represent convolution kernels for image processing operations. They are created by a number of class methods that generate common kernels such as Gaussian, Laplacian, etc.

ImageCore.Image machinevisiontoolbox.ImageCore.Image machinevisiontoolbox.Image

Seealso:

Gauss Laplace Sobel DoG LoG DGauss Circle Box

Methods

Box(h[, normalize])	Square structuring element
Circle(radius[, h, normalize, dtype])	Circular structuring element
DGauss(sigma[, h])	Derivative of Gaussian kernel
disp3d([block])	Show kernel as a 3D surface plot
DoG(sigma1[, sigma2, h])	Difference of Gaussians kernel
Gauss(sigma[, h])	Gaussian kernel
HGauss(sigma[, h])	Hessian of Gaussian kernel
Laplace()	Laplacian kernel
LoG(sigma[, h])	Laplacian of Gaussian kernel
print([fmt, separator, precision])	Print kernel weights in compact format
Sobel()	Sobel edge detector

Properties

shape
T