Module category: Filter

• Register
  • Parameters
  • Input Ports
  • Output Ports
  • Examples

Register

The Registration module can registrate an image stack. It was created for processing together with the ReadITK module.

The module owns four ports for incoming and four ports for outgoing data. The first one delivers an uniform grid, the three following ports the RGB data.

The registration of an image stack can be modified by a few parameters:

Optimizer

You can choose the optimizertype out of a list. It is used to optimize the values of the metric. You can choose between:

• itk::RegularStepGradientDescentOptimizer
• itk::GradientDescentOptimizer
• itk::OnePlusOneEvolutionaryOptimizer

MaxStepLength, MinStepLength, MinStepLengthDiv

The parameters of the optimizer modify the minimal and maximal step length, which are used for the search for the optimal metric value.

Metric

The metric quantitatively measures how well the images fits by comparing the gray-scale intensity of the images. You can choose between:

• itk::MeanSquaresImageToImageMetric
• itk::MattesMutualInformationImageToImageMetric

The MeanSquaresImageToImageMetric computes the mean squared pixel-wise difference in intensity between the images. It relies on the assumption that intensity representing the same homologous point must be the same in both images. Hence, its use is restricted to images of the same modality. The MattesMutualInformationImageToImageMetric works with only one set of intensity samples, which is drawn from the images. This way it works much more faster.

FillColor

The FillColor can be chosen between values of 0 (black) to 255 (white). With this color empty voxels are filled up, in case of transformation and/or rotation cause empty voxels.

Iterations

The amount of the iterations of the registration can be restricted by this value.

Pyramid

You can choose the amount of levels for the image pyramid in addition to speed up the registration. With every level, the images are scaled to the half size based on the image pyramid level before.

Parameters

Name	Type	Description
Optimizer	Choice	Optimizer type
MaxStepLength	FloatScalar	Maximum Step Length for Optimizer
MinStepLength	FloatScalar	Minimum Step Length for Optimizer
MinStepLengthDiv	FloatScalar	Minimum Step Length Divisor for higher Levels
Metric	Choice	Metric type
FillColor	IntScalar	Gray scale fill color for background of transformated images
Iterations	IntScalar	Number of Iterations
Pyramid	IntScalar	Number of Pyramidlevels

Input Ports

Name	Type(s)	Description
VolumeGridIn	UniformGrid	Grid
redin	Float	RGB - Red
greenin	Float	RGB - Green
bluein	Float	RGB - Blue

Output Ports

Name	Type(s)	Description
VolumeGridOut	UniformGrid	Grid
redout	Float	RGB - Red
greenout	Float	RGB - Green
blueout	Float	RGB - Blue

Examples

A "classical" combination for this module is the example Registration.net. A series of images is read in by ReadITK and aligned by Register. While the grid is directly lead to a collect module, an alpha channel is added to the RGB data by the module ColorDistance. After modification by the module Scalar2Vector, the data is lead to the Collect module, which delivers the volume data to a renderer.