Module: Deriche_Filter

Description:

Performs a convolution on the image by a Gaussian with a given approximation order, sigma (can be different for each axis) and derivative order (0 for a blurring, 1 for a gradient, 2 for a Laplacian). For more information, see Recursively implementing the gaussian and its derivatives by R. Deriche (Technical Report 1893, INRIA, May 1993).

Example image.

From left to right: convolution with a Gaussian of sigma = 5, derivative of gaussian of sigma = 2, and a Laplacian of Gaussian of sigma = 2.

Convolution with a Gaussian with a higher sigma for a given axis. Sigma vectors from left to right: (10,1,1), (1,10,1) and (1,1,10).

Connections:

Image

[required]
From this port, the input image is retrieved. The input type must be HxUniformScalarField3.

Ports:

Anisotropic

Tick this box to apply a different sigma for each axis.

Sigma

Value of sigma.

SigmaX

If anisotropic blur is ticked, then this is the sigma used in the X direction.

SigmaY

If anisotropic blur is ticked, then this is the sigma used in the Y direction.

SigmaZ

If anisotropic blur is ticked, then this is the sigma used in the Z direction.

Order

The approximation order of the Gaussian, between 2 and 4. See cited article for details.

Derivative

Order of the derivative: 0 for a blurring, 1 for a gradient, and 2 for a Laplacian.

Anisotropic

For the gradient, output type can be a either the norm of the gradient, or the vector.

HAYAAAAI

Pressing this button starts the computation.

Commands:

Additional options can be accessed when typing in the console Deriche_Filter COMMAND_NAME.

`verbose`

Displays timing information after the computation. Retype to hide info.

`create`

Runs the computation. Returns the name of the output, so it can be used in a script, such as set RESULT [Deriche_Filter create].