Module kernels

Common convolution kernels and traits to aid in the building of kernels

Structs

BoxLinearFilter

The box linear filter is roughly defined as 1/(R*C)*Array2::ones((R, C)) This filter will be a box linear for every colour channel provided

GaussianFilter

Builds a Gaussian kernel taking the covariance as a parameter. Covariance is given as 2 values for the x and y variance.

LaplaceFilter

Create a Laplacian filter, this provides the 2nd spatial derivative of an image.

SobelFilter

Builder to create either a horizontal or vertical Sobel filter for the Sobel operator

Enums

LaplaceType

Specifies the type of Laplacian filter

Orientation

Orientation of the filter

Traits

FixedDimensionKernelBuilder

Create a kernel with a fixed dimension

KernelBuilder

Builds a convolutioon kernel given a shape and optional parameters