Type specific implementations of the interface class for clustering algorithms. More...

#include <mlClusterAlgorithm.h>

Inheritance diagram for ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >:

Public Types
typedef CLUSTERVOXELTYPE	ClusterVoxelType
	The used type of the voxels, needed by derived classes.

Public Member Functions
	ClusterAlgorithm (const ComputeClusterParameters &parameters, Clusters *clusters)
	Constructor to be used.

MLErrorCode	runClustering ()

MLErrorCode	processSlice (MLint zi, const ImageVector &inputImageExtent, TSubImage< MLint8 > &inMaskSlice)

Public Member Functions inherited from ml::ClusterAlgorithmBase
	ClusterAlgorithmBase (const ComputeClusterParameters &parameters, Clusters *clusters)
	Constructor to be used.

virtual	~ClusterAlgorithmBase ()
	Default virtual destructor.

bool	hasResults () const
	Returns whether getCluster() and getClusterOfVoxel() contains cluster data.

Protected Member Functions
MLuint	updateClusterRefForNeighbor (void *neighborVoxelPtr, MLint x, MLint y, MLint z, CLUSTERVOXELTYPE currentValue, MLuint currentClusterRef)
	Checks if neighbor at position x,y,z belongs to the same cluster given by currentClusterRef.

MLErrorCode	calcClustersForSlices ()
	Template applying cluster calculations for one slice.

Protected Member Functions inherited from ml::ClusterAlgorithmBase
void	setInSliceVoxelType (MLDataType dt)
	Sets voxel type of input slices.

void	freeInSliceData ()
	Releases memory of input image slices.

void	storeNextSlice (SubImage &slice)
	Sets next image data, expects the data to be managed by memory manager.

MLErrorCode	loadNextSlice (MLint sliceNr, const ImageVector &inImgExt)
	Gets next image data.

MLErrorCode	loadMaskSlice (TSubImage< MLint8 > &slice, MLint sliceNr, const ImageVector &inImgExt) const
	Gets current slice of the mask image.

MLuint	mergeClusterReference (MLint x, MLint y, MLint z, MLuint currClusterRef)
	Merges a cluster reference currClusterRef for position (x,y,z).

void	setResultFlag (bool result)
	Sets internal result flag to result.

Static Protected Member Functions
static CLUSTERVOXELTYPE	convertBackgroundValue (MLdouble backgroundValue)

Additional Inherited Members
Protected Types inherited from ml::ClusterAlgorithmBase
enum	{ _NR_OF_SUBIMAGES = 2 }
	We do not load the whole image into memory, only keep two slices at once. More...

Protected Attributes inherited from ml::ClusterAlgorithmBase
enum ml::ClusterAlgorithmBase:: { ... }	NumInImagesEnum
	We do not load the whole image into memory, only keep two slices at once.

Clusters *	_clusters
	Result clusters.

SubImage	_currentSlices [_NR_OF_SUBIMAGES]
	do not load the whole image into memory, only keep two slices at once

ComputeClusterParameters	_parameters

Detailed Description

template<typename CLUSTERVOXELTYPE, typename DerivedAlgorithm>
class ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >

Type specific implementations of the interface class for clustering algorithms.

Derived classes should derive from this class.

The derived class needs to pass its class name as second template argument to ClusterAlgorithm, this allows static dispatch to overloaded members (Curiously recurring template pattern).

The derived class needs to implement getVoxel() and isInTolerance(), see IdenticalIntensitiesClusterAlgorithm for an example.

Definition at line 221 of file mlClusterAlgorithm.h.

Member Typedef Documentation

◆ ClusterVoxelType

template<typename CLUSTERVOXELTYPE , typename DerivedAlgorithm >

typedef CLUSTERVOXELTYPE ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >::ClusterVoxelType

The used type of the voxels, needed by derived classes.

Definition at line 226 of file mlClusterAlgorithm.h.

Constructor & Destructor Documentation

◆ ClusterAlgorithm()

template<typename CLUSTERVOXELTYPE , typename DerivedAlgorithm >

ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >::ClusterAlgorithm	(	const ComputeClusterParameters &	parameters,
		Clusters *	clusters )

inline

Constructor to be used.

Definition at line 229 of file mlClusterAlgorithm.h.

Member Function Documentation

◆ calcClustersForSlices()

template<typename CLUSTERVOXELTYPE , typename DerivedAlgorithm >

MLErrorCode ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >::calcClustersForSlices ( )

inlineprotected

Template applying cluster calculations for one slice.

Returns: ML_RESULT_OK in case of successful processing, otherwise an error code describing the problem.

Definition at line 269 of file mlClusterAlgorithm.h.

References ml::ClusterHandler< CLUSTERVOXELTYPE, DerivedAlgorithm >::getErrorCode(), ML_RESULT_OK, mlrange_cast(), NBH_2D_4_XY, NBH_2D_8_XY, ml::ClusterHandler< CLUSTERVOXELTYPE, DerivedAlgorithm >::setParameters(), and ml::ClusterUserDataParameters::useImageValueAsUserData.

◆ convertBackgroundValue()

template<typename CLUSTERVOXELTYPE , typename DerivedAlgorithm >

static CLUSTERVOXELTYPE ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >::convertBackgroundValue ( MLdouble backgroundValue )

inlinestaticprotected

Definition at line 262 of file mlClusterAlgorithm.h.

References mlrange_cast().

◆ processSlice()

template<typename CLUSTERVOXELTYPE , typename DerivedAlgorithm >

MLErrorCode ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >::processSlice	(	MLint	zi,
		const ImageVector &	inputImageExtent,
		TSubImage< MLint8 > &	inMaskSlice )

inline

Definition at line 345 of file mlClusterAlgorithm.h.

References ML_PROGRAMMING_ERROR, ML_RESULT_OK, mlrange_cast(), MLSizeOf(), NBH_2D_4_XY, NBH_2D_8_XY, NBH_3D_18_XYZ, NBH_3D_26_XYZ, NBH_3D_6_XYZ, and NBH_COUNT.

◆ runClustering()

template<typename CLUSTERVOXELTYPE , typename DerivedAlgorithm >

MLErrorCode ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >::runClustering ( )

inline

Definition at line 233 of file mlClusterAlgorithm.h.

References mlrange_cast().

◆ updateClusterRefForNeighbor()

template<typename CLUSTERVOXELTYPE , typename DerivedAlgorithm >

MLuint ml::ClusterAlgorithm< CLUSTERVOXELTYPE, DerivedAlgorithm >::updateClusterRefForNeighbor	(	void *	neighborVoxelPtr,
		MLint	x,
		MLint	y,
		MLint	z,
		CLUSTERVOXELTYPE	currentValue,
		MLuint	currentClusterRef )

inlineprotected

Checks if neighbor at position x,y,z belongs to the same cluster given by currentClusterRef.

Parameters

neighborVoxelPtr	Pointer to the neighbor voxel
x,y,z	Position of neighbor in image coordinates.
currentValue	Currently examined voxel.
currentClusterRef	Cluster reference of currently examined voxel.

Returns: The cluster reference after checking the neighbor. Can differ from currentClusterRef due to cluster merging.

Definition at line 248 of file mlClusterAlgorithm.h.

References mlrange_cast().

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

MeVisLab/Standard/Sources/Shared/MLClusterAlgorithm/mlClusterAlgorithm.h

Public Types

Public Member Functions

Protected Member Functions

Static Protected Member Functions

Additional Inherited Members

Detailed Description

Member Typedef Documentation

◆ ClusterVoxelType

Constructor & Destructor Documentation

◆ ClusterAlgorithm()

Member Function Documentation

◆ calcClustersForSlices()

◆ convertBackgroundValue()

◆ processSlice()

◆ runClustering()

◆ updateClusterRefForNeighbor()