Provides static methods for image transformations including multi-resolution decomposition, Laplacian of Gaussian filtering, wavelet transforms, and image resampling. More...

#include <mitkTransformationOperation.h>

Static Public Member Functions
static std::vector< Image::Pointer >	MultiResolution (Image::Pointer &image, unsigned int numberOfLevels, bool outputAsDouble=false)
	Compute a multi-resolution Gaussian pyramid decomposition of an image. More...

static Image::Pointer	LaplacianOfGaussian (Image::Pointer &image, double sigma, bool outputAsDouble=false)
	Apply a Laplacian of Gaussian (LoG) filter to an image. More...

static std::vector< Image::Pointer >	WaveletForward (Image::Pointer &image, unsigned int numberOfLevels, unsigned int numberOfBands, BorderCondition condition, WaveletType waveletType)
	Compute a forward wavelet decomposition of an image. More...

static Image::Pointer	ResampleImage (Image::Pointer &image, mitk::Vector3D spacing, mitk::ImageMappingInterpolator::Type interpolator, GridInterpolationPositionType position, bool returnAsDouble, bool roundOutput)
	Resample an image to a new voxel spacing. More...

static Image::Pointer	ResampleMask (Image::Pointer &image, mitk::Vector3D spacing, mitk::ImageMappingInterpolator::Type interpolator, GridInterpolationPositionType position)
	Resample a binary mask image to a new voxel spacing. More...

Detailed Description

Provides static methods for image transformations including multi-resolution decomposition, Laplacian of Gaussian filtering, wavelet transforms, and image resampling.

The TransformationOperation class offers a set of static methods for spatial and frequency-domain transformations on mitk::Image objects. These include:

Multi-resolution Gaussian pyramid decomposition
Laplacian of Gaussian edge detection
Forward wavelet decomposition with selectable wavelet families
Image resampling with configurable interpolation and grid positioning
Mask resampling with automatic thresholding for non-nearest-neighbor interpolators

See also: ArithmeticOperation, MaskCleaningOperation

Definition at line 71 of file mitkTransformationOperation.h.

Member Function Documentation

◆ LaplacianOfGaussian()

static Image::Pointer mitk::TransformationOperation::LaplacianOfGaussian	(	Image::Pointer &	image,
		double	sigma,
		bool	outputAsDouble = `false`
	)

static

Apply a Laplacian of Gaussian (LoG) filter to an image.

Computes the Laplacian of the image after Gaussian smoothing with the specified sigma. This is useful for edge detection and blob detection at a given scale.

Parameters

[in]	image	The input image to filter.
[in]	sigma	The standard deviation of the Gaussian kernel in physical units.
[in]	outputAsDouble	If true, the output image has double pixel type. Default: false.

Returns: A new image containing the Laplacian of Gaussian filtered result.

◆ MultiResolution()

static std::vector<Image::Pointer> mitk::TransformationOperation::MultiResolution	(	Image::Pointer &	image,
		unsigned int	numberOfLevels,
		bool	outputAsDouble = `false`
	)

static

Compute a multi-resolution Gaussian pyramid decomposition of an image.

Uses ITK's RecursiveMultiResolutionPyramidImageFilter to produce a set of progressively lower-resolution versions of the input image. Level 0 is the lowest resolution.

Parameters

[in]	image	The input image to decompose.
[in]	numberOfLevels	The number of pyramid levels to generate.
[in]	outputAsDouble	If true, output images have double pixel type. Default: false.

Returns: A vector of images, one per pyramid level, ordered from lowest to highest resolution.

◆ ResampleImage()

static Image::Pointer mitk::TransformationOperation::ResampleImage	(	Image::Pointer &	image,
		mitk::Vector3D	spacing,
		mitk::ImageMappingInterpolator::Type	interpolator,
		GridInterpolationPositionType	position,
		bool	returnAsDouble,
		bool	roundOutput
	)

static

Resample an image to a new voxel spacing.

Resamples the input image to the specified spacing using the given interpolation method and grid positioning strategy. Only spacing components greater than 0 are applied; components set to 0 or less retain the original spacing.

Parameters

[in]	image	The input image to resample.
[in]	spacing	The desired output spacing vector. Components <= 0 are ignored.
[in]	interpolator	The interpolation method to use (e.g., Linear, NearestNeighbor, BSpline3).
[in]	position	The grid positioning strategy controlling how the output grid aligns with the input image.
[in]	returnAsDouble	If true, the image is first cast to double pixel type before resampling.
[in]	roundOutput	If true, the resampled result is rounded back to the original pixel type.

Returns: A new image resampled to the specified spacing.

See also: GridInterpolationPositionType, ResampleMask

◆ ResampleMask()

static Image::Pointer mitk::TransformationOperation::ResampleMask	(	Image::Pointer &	image,
		mitk::Vector3D	spacing,
		mitk::ImageMappingInterpolator::Type	interpolator,
		GridInterpolationPositionType	position
	)

static

Resample a binary mask image to a new voxel spacing.

Specialized resampling for binary mask images. When using nearest-neighbor interpolation, this delegates directly to ResampleImage(). For other interpolation methods, the mask is resampled as a double image and then thresholded at 0.5 to produce a binary result, preserving mask boundaries more accurately than direct nearest-neighbor resampling.

Parameters

[in]	image	The input mask image to resample (unsigned short expected).
[in]	spacing	The desired output spacing vector. Components <= 0 are ignored.
[in]	interpolator	The interpolation method to use.
[in]	position	The grid positioning strategy.

Returns: A new binary mask image resampled to the specified spacing.

See also: ResampleImage, GridInterpolationPositionType

◆ WaveletForward()

static std::vector<Image::Pointer> mitk::TransformationOperation::WaveletForward	(	Image::Pointer &	image,
		unsigned int	numberOfLevels,
		unsigned int	numberOfBands,
		BorderCondition	condition,
		WaveletType	waveletType
	)

static

Compute a forward wavelet decomposition of an image.

Performs a frequency-domain wavelet decomposition using the specified isotropic wavelet function. The input image is FFT-padded with the chosen boundary condition before transformation.

Parameters

[in]	image	The input image to decompose.
[in]	numberOfLevels	The number of decomposition levels.
[in]	numberOfBands	The number of high-pass sub-bands per level.
[in]	condition	The boundary condition for FFT padding.
[in]	waveletType	The isotropic wavelet function to use.

Returns: A vector of images representing the wavelet sub-bands. For each level, numberOfBands images are produced, plus one low-pass image at the coarsest level.

See also: BorderCondition, WaveletType

The documentation for this class was generated from the following file:

mitkTransformationOperation.h

Static Public Member Functions

Detailed Description

Member Function Documentation

◆ LaplacianOfGaussian()

◆ MultiResolution()

◆ ResampleImage()

◆ ResampleMask()

◆ WaveletForward()