This class manages/represents a rectangular 6d image region which is organized linearly in memory. More...

#include <mlSubImage.h>

Inheritance diagram for ml::SubImage:

Public Member Functions
const SubImageBox &	getBox () const
	Returns the box describing the origin/extent of the subimage.

void	setSourceImageExtent (const ImageVector &extent)
	Sets the image extent of the source image that was used to create this SubImage to `extent`.

ImageVector	getSourceImageExtent () const
	Returns the image extent of the source image that was used to create this SubImage.

SubImageBox	getValidRegion () const
	Returns the valid region of the SubImage, which is defined by the intersection of the SubImage::getBox() and the SubImageBox(ImageVector(0), SubImage::getSourceImageExtent()-1).

MLint	getNumVoxels () const
	Returns number of voxels in (sub)image.

MLint	getSizeInBytes () const
	Returns number of potential bytes in (sub)image.

ImageVector	getStride () const
	Returns a stride vector to address the memory efficiently.

MLint	getOffset (const ImageVector &voxelPosition) const
	Returns the array index offset from the origin (0,0,0,0,0,0) to the `voxelPosition`.

void	setDataType (MLDataType dataType)
	Overwrite inherited setDataType method to set type of data to `dataType`.

MLDataType	getDataType () const
	Return type of image data.

const MLTypeInfos *	getDataTypeInfos () const
	Get MLTypeInfos for image data type.

MLEXPORT ImageProperties	toImageProperties () const
	Converts the Subimage's datatype and extend to an ImageProperties object.

MLEXPORT void	setFromImageProperties (const ImageProperties &imageProperties)
	Sets the image box and datatype from the `imageProperties` extent and datatype.

void	toStream (std::ostream &ostr) const

Constructors and assignments.
	SubImage ()
	Constructor: Creates a subimage with no data.

	SubImage (const SubImage &si)
	Copy constructor to get identical copy.

	SubImage (const SubImageBox &box, MLDataType datatype, void *data=nullptr)
	Constructor for a rectangular 6d image region with position and extent given by box, with data type datatype and memory given by the pointer data.

	SubImage (const SubImage &si, const ImageVector &offset)
	Copy constructor to get identical copy.

virtual	~SubImage ()=default
	Virtual destructor to suppress compiler warnings.

SubImage &	operator= (const SubImage &si)
	Assignment operator to get identical copy.

Origin, extents, region of image, offsets and strides.
void	setBox (const SubImageBox &subImageBox)
	Sets a rectangular 6d region of the subimage to `subImageBox`.

void	setBox (const ImageVector &imageExtent)
	Sets a rectangular 6d region of the subimage to `imageExtent`.

void	translate (const ImageVector &offset)
	Translates the box of the subimage by adding the vector `offset`.

const ImageVector &	getOrigin () const
	Returns the origin of the subimage (which is identical to getBox().v1).

void	setOrigin (const ImageVector &newOrigin)
	Sets the origin of the subimage (which moves the box of the subimage to `newOrigin`).

ImageVector	getExtent () const
	Returns the extent of the subimage (which is identical to getBox().getExtent()).

void	setExtent (MLint x, MLint y, MLint z=1, MLint c=1, MLint t=1, MLint u=1)
	Sets the extent of the subimage (which also changes the strides) to `newExtent`.

void	setExtent (const ImageVector &newExtent)
	Sets the extent of the subimage (which also changes the strides) to `newExtent`.

SubImageBox	getBoxFromExtent () const
	Returns size of image as box with origin 0.

Deprecated
Deprecated Use getBoxFromExtent() instead. Returns extent of the image as box with origin 0.
SubImageBox	getBoxFromImageExtent () const

ImageVector	getImageExtent () const

void	setImageExtent (const ImageVector &newExtent)

Calculation of memory addresses of image voxels
void *	getSubImagePointer (const ImageVector &voxelPosition) const
	Returns pointer to voxel data of image voxel at 6d `voxelPosition` relative to the begin of the subimage region.

void *	getSubImagePointer (MLint x, MLint y, MLint z) const
	Returns pointer to voxel data of image voxel at 3d position p=(`x`,`y`,`z`) relative to the begin of the subimage region.

void *	getImagePointer (const ImageVector &voxelPosition) const
	Returns pointer to voxel data of image voxel at 6d position `voxelPosition` relative to the begin of the complete image region.

void *	getImagePointer (MLint x, MLint y, MLint z) const
	Returns pointer to voxel data of image voxel at 3d position p=(`x`, `y`, `z`) relative to the begin of the complete image region.

Copy image data from one subimage to another.
MLEXPORT void	copySubImage (const SubImage &fromImage, const ScaleShiftData &scaleShiftData)
	Copies image data from the subimage `fromImage` into the overlapping region of this subimage.

MLEXPORT void	copySubImage (const SubImage &fromImage)
	Implements special case for void copySubImage(const TSubImage<FROM_DATATYPE> &typedFromImg, const ScaleShiftData& scaleShiftData) where scaleShiftData is default.

Test for validity of coordinates.
bool	isValidSubImagePosition (const ImageVector &voxelPosition) const
	Returns `true` if 6d `voxelPosition` is a valid position within subimage region, i.e., if `voxelPosition` is within (0,0,0,0,0,0) and subimage extents.

bool	isValidSubImagePosition (MLint x, MLint y, MLint z) const
	Returns `true` if 3d position p=(`x`, `y`, `z`) is a valid position within subimage region, i.e., if the position is within (0,0,0) and subimage extents.

bool	isValidImagePosition (const ImageVector &voxelPosition) const
	Returns `true` if 6d `voxelPosition` is a valid position within subimage region, i.e., if `voxelPosition` is within subImgBox.v1 and subImgBox.v2.

bool	isValidImagePosition (MLint x, MLint y, MLint z) const
	Returns `true` if 3d position p=(`x`, `y`, `z`) is a valid position within subimage region, i.e., if the position is within subImgBox.v1 and subImgBox.v2.

Allocation and freeing of buffered data.
MLEXPORT void	allocate (MLMemoryErrorHandling handleFailure)
	Allocate data with MLAlloc().

MLEXPORT void	allocateAsMemoryBlockHandle (MLMemoryErrorHandling handleFailure=ML_RETURN_NULL)
	Allocate data using the ML memory manager. For failure handing, see SubImage::allocate().

MLEXPORT void	free ()
	Frees data pointed to by getData() with MLFree().

Special properties
MLEXPORT bool	isOneValued () const
	Returns `true` if all page voxels have the same value or if the page contains only 1 voxel.

MLEXPORT MLint	calculateMinMax (MLdouble &minValue, MLdouble &maxValue, const SubImageBox *const validBox=nullptr) const
	Scans subimage and determines minimum and maximum values on region which is part of subimage box and of validBox.

MLEXPORT void	compare (const SubImage &subImage2, bool regionsMatch, bool dataTypesMatch, bool thisBoxIsPartOfRegion2, bool region2IsPartOfThisBox, bool overlapHasSameValues, ImageVector firstMismatchPos) const
	Compare two subimages with respect to their regions and their contents.

SUBIMAGE FILLING
MLEXPORT void	fill (MLdouble value)
	Fills the subimage with a `value` cast to the data type of the subimage.

MLEXPORT void	fillWithTypeData (const MLTypeData *value)
	Fills the subimage with a `value` given as MLTypeData.

MLEXPORT void	fillBordersWithScalarValue (const SubImageBox &box, MLdouble fillValue)
	Fill all regions with the fill value fillValue which are not covered by box.

MLEXPORT void	fillBordersWithTypeData (const SubImageBox &box, const MLTypeData *fillValue)
	Fills all regions with the fill value `fillValue` which are not covered by `box`.

MLEXPORT void	fillBordersWithInputValues (const SubImageBox &box, const SubImage &inputSubImage)
	Fills all regions with values from `inputSubImage` which are not covered by `box`, however, only valid regions of `inputSubImage` are read.

MLEXPORT void	fillBordersWithBorderValues (const SubImageBox &box)
	Each voxel of which is not inside `box` is filled with the nearest voxel which is inside `box`.

void	fillInvalidRegionWithScalarValue (MLdouble value)
	Fills the invalid region (everything outside of getValidRegion()) with the given `value`.

void	fillInvalidRegionWithTypeData (const MLTypeData *value)
	Fills the invalid region (everything outside of getValidRegion()) with the given `value`.

void	fillInvalidRegionWithBorderValues ()
	Fills the invalid region (everything outside of getValidRegion()) with the values on the borders of the valid region.

Protected Member Functions
void	_calcFillAreaParams (const SubImageBox &box, const SubImageBox &maxValidInputRegion, ImageVector &boxV1, ImageVector &boxV2, ImageVector &outputTSubImageV1, ImageVector &outputTSubImageV2, MLint &fullLineLenX, MLint &fullLineLenXB, MLint &leftLineStartX, MLint &leftLineLenX, MLint &leftLineLenXB, MLint &rightLineStartX, MLint &rightLineLenX, MLint &rightLineLenXB)
	Helper method to calculate important parameters for page based border filling.

void	_copySubImageGeneric (const SubImage &fromImage, const ScaleShiftData &scaleShiftData)
	Generic, non-optimized version of copySubImage.

const MLTypeInfos *	_getDataTypeInfos (MLDataType dt) const
	Get MLTypeInfos for used data types, indicates error if type is not registered.

Protected Attributes
SubImageBox	_box
	The box of the subimage.

ImageVector	_sourceImageExtent
	The extent of the source image, which is used for getValidRegion()

void *	_data
	Memory chunk managed by this subimage.

MLMemoryBlockHandle	_memoryBlock
	Memory block used by this subimage.

ImageVector	_stride
	Stride vector to address the memory efficiently.

MLDataType	_dataType
	Datatype of the image.

Properties of subimage
MLEXPORT bool	isValid () const
	Returns 1(=true) if image region and data is valid.

void *	getData () const
	Returns the memory address of the memory managed by the subimage.

MLEXPORT void	setData (void *data)
	Sets `data` as a new memory block for the subimage.

MLEXPORT void	setDataFromMemoryBlockHandle (const MLMemoryBlockHandle &data)
	Sets the managed `data` from a given MLMemoryBlockHandle, the sub image will store this handle and thus reference count the memory managed by the handle as long as the SubImage stays alive.

const MLMemoryBlockHandle &	getMemoryBlockHandle () const
	Returns the memory block handle that manages the data of this SubImage if it has been set via setDataFromMemoryBlockHandle().

static MLEXPORT MLint	coordToIndex (MLint x, MLint y, MLint z, MLint c, MLint t, MLint u, const ImageVector &size)
	Converts the coordinate (`x`, `y`, `z`, `c`, `t`,`u`) to an index into an image with 6D extents given by `size`.

static MLEXPORT MLint	coordToIndex (const ImageVector &voxelPosition, const ImageVector &size)
	Converts the coordinate `voxelPosition` into the image with extents `size` to an index.

static MLEXPORT ImageVector	indexToCoord (MLint index, const ImageVector &extent)
	Converts an `index` into an array with extents `extent` to a coordinate.

Detailed Description

This class manages/represents a rectangular 6d image region which is organized linearly in memory.

Thread-safety: This class is reentrant.

This class owns

Methods to set/access image properties which are inherited from the class ImageProperties.
A pointer to the memory data containing the image data as void pointer.

IMPORTANT NOTES:

This subimage class does NOT allocate/deallocate the memory chunk itself. It usually must be allocated elsewhere and assigned to this class.
The class itself is not exported to avoid that inline usage is inhibited. Only stuff implemented in the .cpp file is exported explicitly.

Other notes:

With this class the Host manages and encapsulates rectangular image regions (e.g., for pages, tiles, cached image results) and passes them to the image processing algorithms.
The typical image processing methods in the ML are located in overloaded methods of Module::calculateOutputSubImage. In these methods usually the untyped memory chunks given as SubImage are wrapped again to typed subimages (see TSubImage). See TSubImage for more information about that.
The type of the image data in memory is handled via a void pointer, however the type is managed as enum type to support typed accesses in derived classes (see TSubImage). Thus the access to certain image voxels is not supported. The Host does not need the knowledge about the data itself. The typed access to voxels is implemented on top of this class as the template class TSubImage.

Definition at line 74 of file mlSubImage.h.

Constructor & Destructor Documentation

◆ SubImage() [1/4]

ml::SubImage::SubImage ( )

inline

Constructor: Creates a subimage with no data.

Definition at line 84 of file mlSubImage.h.

◆ SubImage() [2/4]

ml::SubImage::SubImage ( const SubImage & si )

inline

Copy constructor to get identical copy.

Note that the pointer to the data is copied, and NOT the data itself.

Definition at line 97 of file mlSubImage.h.

◆ SubImage() [3/4]

ml::SubImage::SubImage	(	const SubImageBox &	box,
		MLDataType	datatype,
		void *	data = nullptr )

inline

Constructor for a rectangular 6d image region with position and extent given by box, with data type datatype and memory given by the pointer data.

Definition at line 111 of file mlSubImage.h.

◆ SubImage() [4/4]

ml::SubImage::SubImage	(	const SubImage &	si,
		const ImageVector &	offset )

inline

Copy constructor to get identical copy.

Note that the pointer to the data is copied, and NOT the data itself. The box of the input image is translated by the given offset vector.

Definition at line 124 of file mlSubImage.h.

◆ ~SubImage()

virtual ml::SubImage::~SubImage ( )

inlinevirtualdefault

Virtual destructor to suppress compiler warnings.

Member Function Documentation

◆ _calcFillAreaParams()

void ml::SubImage::_calcFillAreaParams	(	const SubImageBox &	box,
		const SubImageBox &	maxValidInputRegion,
		ImageVector &	boxV1,
		ImageVector &	boxV2,
		ImageVector &	outputTSubImageV1,
		ImageVector &	outputTSubImageV2,
		MLint &	fullLineLenX,
		MLint &	fullLineLenXB,
		MLint &	leftLineStartX,
		MLint &	leftLineLenX,
		MLint &	leftLineLenXB,
		MLint &	rightLineStartX,
		MLint &	rightLineLenX,
		MLint &	rightLineLenXB )

protected

Helper method to calculate important parameters for page based border filling.

Input parameters: box specifies the box which shall not be filled. maxValidInputRegion is the region which is allowed to be read.

Return parameters: boxV1 and boxV2 are the corners of box. outputTSubImageV1 and outputTSubImageV2 are the corners of the box of this. fullLineLenX is the number of voxels in one row of this. fullLineLenXB is the number of bytes in one row of this. leftLineStartX is the first voxel left from box to be filled in x-rows. leftLineLenX is the number of voxels to be filled starting at voxel leftLineStartX. leftLineLenXB is the number of bytes to be filled starting at voxel leftLineStartX. rightLineStartX is the first voxel right from box to be filled in x-rows. rightLineLenX is the number of voxels to be filled starting at voxel rightLineStartX. rightLineLenXB is the number of bytes to be filled starting at voxel rightLineStartX.

◆ _copySubImageGeneric()

void ml::SubImage::_copySubImageGeneric	(	const SubImage &	fromImage,
		const ScaleShiftData &	scaleShiftData )

protected

Generic, non-optimized version of copySubImage.

◆ _getDataTypeInfos()

const MLTypeInfos * ml::SubImage::_getDataTypeInfos ( MLDataType dt ) const

protected

Get MLTypeInfos for used data types, indicates error if type is not registered.

◆ allocate()

MLEXPORT void ml::SubImage::allocate ( MLMemoryErrorHandling handleFailure )

Allocate data with MLAlloc().

If the allocation fails it is handled as described by handleFailure (see MLMemoryErrorHandling for description of modes). The internal pointer to the data buffer is simply overwritten; the application is responsible for freeing it before, for example with SubImage::free(). The number of allocated bytes results from the size of the subimage box and the data type. An empty box will result in the allocation of zero bytes on the heap which also need to be freed correctly.

IMPORTANT: By definition the allocated data is not owned by the subimage and will NOT be freed on destruction of the SubImage instance. The calling application is responsible for that. Use SubImage::free() for that.

◆ allocateAsMemoryBlockHandle()

MLEXPORT void ml::SubImage::allocateAsMemoryBlockHandle ( MLMemoryErrorHandling handleFailure = ML_RETURN_NULL )

Allocate data using the ML memory manager. For failure handing, see SubImage::allocate().

◆ calculateMinMax()

MLEXPORT MLint ml::SubImage::calculateMinMax	(	MLdouble &	minValue,
		MLdouble &	maxValue,
		const SubImageBox *const	validBox = nullptr ) const

Scans subimage and determines minimum and maximum values on region which is part of subimage box and of validBox.

If validBox is passed as a NULL pointer (the default), all voxels are checked. The return value is the number of scanned voxels. If the subimage is empty, minValue and maxValue are set to 0 and the return value is 0. Warning: If used in calculateOutputSubImage, take into consideration that pages may also contain invalid data which is not part of the image. Pass 'image box' as validBox to be sure to access only valid regions.

◆ compare()

MLEXPORT void ml::SubImage::compare	(	const SubImage &	subImage2,
		bool *	regionsMatch,
		bool *	dataTypesMatch,
		bool *	thisBoxIsPartOfRegion2,
		bool *	region2IsPartOfThisBox,
		bool *	overlapHasSameValues,
		ImageVector *	firstMismatchPos ) const

Compare two subimages with respect to their regions and their contents.

All pointer parameters may be passed as NULL parameters if not needed and will silently be ignored then.

Parameters

subImage2	Second subimage to be compared with.
regionsMatch	If passed as non-NULL pointer, *regionsMatch is set to `true` if regions are identical.
dataTypesMatch	If passed as non-NULL pointer, *dataTypesMatch is set to `true` if both data types are identical.
thisBoxIsPartOfRegion2	If passed as non-NULL pointer, *region1IsPartOfRegion2 is set to `true` if the region of this is a subregion of `subImage2` or is identical.
region2IsPartOfThisBox	If passed as non-NULL pointer, *region2IsPartOfRegion1 is set to `true` if the region of `subImage2` is a subregion of this or is identical.
overlapHasSameValues	If passed as non-NULL pointer, *overlapHasSameValues is set to `true` if subimages have the same and valid types, data pointers are both non-NULL and all voxels in the overlapping region of this and `subImage2` contain equal values.
firstMismatchPos	If passed as non-NULL pointer, *firstMismatchPos is set to the first value which does not match in overlapping regions. If no overlap exists or something is invalid (data pointers or types) it is set to (-1, ..., -1).

◆ coordToIndex() [1/2]

static MLEXPORT MLint ml::SubImage::coordToIndex	(	const ImageVector &	voxelPosition,
		const ImageVector &	size )

static

Converts the coordinate voxelPosition into the image with extents size to an index.

This convenience function is static and can be used also without a class instance.

◆ coordToIndex() [2/2]

static MLEXPORT MLint ml::SubImage::coordToIndex	(	MLint	x,
		MLint	y,
		MLint	z,
		MLint	c,
		MLint	t,
		MLint	u,
		const ImageVector &	size )

static

Converts the coordinate (x, y, z, c, t,u) to an index into an image with 6D extents given by size.

This convenience function is static and can be used also without a class instance.

◆ copySubImage() [1/2]

MLEXPORT void ml::SubImage::copySubImage ( const SubImage & fromImage )

Implements special case for void copySubImage(const TSubImage<FROM_DATATYPE> &typedFromImg, const ScaleShiftData& scaleShiftData) where scaleShiftData is default.

◆ copySubImage() [2/2]

MLEXPORT void ml::SubImage::copySubImage	(	const SubImage &	fromImage,
		const ScaleShiftData &	scaleShiftData )

Copies image data from the subimage fromImage into the overlapping region of this subimage.

Grey values are transformed linearly according to the settings in scaleShiftData (see ScaleShiftData). Notes:

The copied data always has the data type of this subimage. If necessary it is cast.
By transforming linearly gray values before copying them it is possible to avoid information loss due to type casting.
The default case of scaleShiftData defines scaleShiftData.getScale()==1 and scaleShiftData.getShift()==0. So voxel value transformation does not change anything and copying is implemented as special case which is therefore faster.

◆ fill()

MLEXPORT void ml::SubImage::fill ( MLdouble value )

Fills the subimage with a value cast to the data type of the subimage.

◆ fillBordersWithBorderValues()

MLEXPORT void ml::SubImage::fillBordersWithBorderValues ( const SubImageBox & box )

Each voxel of which is not inside box is filled with the nearest voxel which is inside box.

Note that box must be specified in global image coordinates. box must overlap the box of the subImage at least in one voxel to guarantee correct operation since copying can be done only from box area. If no overlap exists then the routine returns without filling anything.

◆ fillBordersWithInputValues()

MLEXPORT void ml::SubImage::fillBordersWithInputValues	(	const SubImageBox &	box,
		const SubImage &	inputSubImage )

Fills all regions with values from inputSubImage which are not covered by box, however, only valid regions of inputSubImage are read.

Not available regions from the input are ignored and may lead to unchanged regions outside box if inputSubImage does not cover all regions outside box. Note that box must be specified in global image coordinates. So two subimages with different origins and same sizes are filled differently, dependent on their overlap with the box in global image coordinates. This method should be called only if this and inputSubImage differ.

◆ fillBordersWithScalarValue()

MLEXPORT void ml::SubImage::fillBordersWithScalarValue	(	const SubImageBox &	box,
		MLdouble	fillValue )

Fill all regions with the fill value fillValue which are not covered by box.

Note that box must be specified in global image coordinates. So two subimages with different origins and same sizes are filled differently, dependent on their overlap with the box in global image coordinates.

◆ fillBordersWithTypeData()

MLEXPORT void ml::SubImage::fillBordersWithTypeData	(	const SubImageBox &	box,
		const MLTypeData *	fillValue )

Fills all regions with the fill value fillValue which are not covered by box.

Note that box must be specified in global image coordinates. So two subimages with different origins and same sizes are filled differently, dependent on their overlap with the box in global image coordinates. fillValue must point to a valid and correctly sized value of the data type of the subimage. If fillValue is NULL, the call is ignored.

◆ fillInvalidRegionWithBorderValues()

void ml::SubImage::fillInvalidRegionWithBorderValues ( )

inline

Fills the invalid region (everything outside of getValidRegion()) with the values on the borders of the valid region.

This is useful to initialize invalid regions of an output/input sub image.

Definition at line 633 of file mlSubImage.h.

◆ fillInvalidRegionWithScalarValue()

void ml::SubImage::fillInvalidRegionWithScalarValue ( MLdouble value )

inline

Fills the invalid region (everything outside of getValidRegion()) with the given value.

This is useful to initialize invalid regions of an output/input sub image.

Definition at line 620 of file mlSubImage.h.

◆ fillInvalidRegionWithTypeData()

void ml::SubImage::fillInvalidRegionWithTypeData ( const MLTypeData * value )

inline

Fills the invalid region (everything outside of getValidRegion()) with the given value.

This is useful to initialize invalid regions of an output/input sub image.

Definition at line 626 of file mlSubImage.h.

◆ fillWithTypeData()

MLEXPORT void ml::SubImage::fillWithTypeData ( const MLTypeData * value )

Fills the subimage with a value given as MLTypeData.

value must point to a valid and correctly sized value of the data type of the subimage. If value is NULL, the call is ignored.

◆ free()

MLEXPORT void ml::SubImage::free ( )

Frees data pointed to by getData() with MLFree().

as described by MLFree(). It is legal to call free() if getData() is NULL. In that case the call is ignored. After this call getData() will return NULL. If the memory has been set via setDataFromMemoryBlockHandle(), free() does NOT delete the memory but just clears the internal memory handle.

◆ getBox()

const SubImageBox & ml::SubImage::getBox ( ) const

inline

Returns the box describing the origin/extent of the subimage.

Definition at line 230 of file mlSubImage.h.

Referenced by ml::TVirtualVolume< DATATYPE >::copyMaskToSubImage(), ml::MLKernelToolsApplyFiltering(), and ml::processDataWithITKFilterAndFillOutSubImg().

◆ getBoxFromExtent()

SubImageBox ml::SubImage::getBoxFromExtent ( ) const

inline

Returns size of image as box with origin 0.

Definition at line 212 of file mlSubImage.h.

◆ getBoxFromImageExtent()

SubImageBox ml::SubImage::getBoxFromImageExtent ( ) const

inline

Definition at line 218 of file mlSubImage.h.

◆ getData()

void * ml::SubImage::getData ( ) const

inline

Returns the memory address of the memory managed by the subimage.

Definition at line 372 of file mlSubImage.h.

◆ getDataType()

MLDataType ml::SubImage::getDataType ( ) const

inline

Return type of image data.

Definition at line 288 of file mlSubImage.h.

◆ getDataTypeInfos()

const MLTypeInfos * ml::SubImage::getDataTypeInfos ( ) const

inline

Get MLTypeInfos for image data type.

Definition at line 291 of file mlSubImage.h.

References MLGetTypeInfosForDataType().

◆ getExtent()

ImageVector ml::SubImage::getExtent ( ) const

inline

Returns the extent of the subimage (which is identical to getBox().getExtent()).

Definition at line 197 of file mlSubImage.h.

◆ getImageExtent()

ImageVector ml::SubImage::getImageExtent ( ) const

inline

Deprecated: Use getExtent() instead. Returns the extent of the image.

Definition at line 222 of file mlSubImage.h.

◆ getImagePointer() [1/2]

void * ml::SubImage::getImagePointer ( const ImageVector & voxelPosition ) const

inline

Returns pointer to voxel data of image voxel at 6d position voxelPosition relative to the begin of the complete image region.

The subimage is a part of the complete image. I.e., the position p=getBox.v1() is considered the origin of the image region/subimage.

Definition at line 323 of file mlSubImage.h.

◆ getImagePointer() [2/2]

void * ml::SubImage::getImagePointer	(	MLint	x,
		MLint	y,
		MLint	z ) const

inline

Returns pointer to voxel data of image voxel at 3d position p=(x, y, z) relative to the begin of the complete image region.

The subimage is a part of the complete image. I.e., the position p=getBox.v1() is considered the origin of the image region/subimage.

Definition at line 331 of file mlSubImage.h.

◆ getMemoryBlockHandle()

const MLMemoryBlockHandle & ml::SubImage::getMemoryBlockHandle ( ) const

inline

Returns the memory block handle that manages the data of this SubImage if it has been set via setDataFromMemoryBlockHandle().

Returns an empty handle when the data has been set with setData().

Definition at line 388 of file mlSubImage.h.

◆ getNumVoxels()

MLint ml::SubImage::getNumVoxels ( ) const

inline

Returns number of voxels in (sub)image.

Note that it could become larger than 2^32 even on 32 bit systems. It will never be smaller than 0; the signed return type is used for safer signed arithmetics.

Definition at line 255 of file mlSubImage.h.

◆ getOffset()

MLint ml::SubImage::getOffset ( const ImageVector & voxelPosition ) const

inline

Returns the array index offset from the origin (0,0,0,0,0,0) to the voxelPosition.

Definition at line 270 of file mlSubImage.h.

◆ getOrigin()

const ImageVector & ml::SubImage::getOrigin ( ) const

inline

Returns the origin of the subimage (which is identical to getBox().v1).

Definition at line 186 of file mlSubImage.h.

◆ getSizeInBytes()

MLint ml::SubImage::getSizeInBytes ( ) const

inline

Returns number of potential bytes in (sub)image.

Note that it could become larger than 2^32 even on 32 bit systems. It will never be smaller than 0; the signed return type is used for safer signed arithmetics.

Definition at line 259 of file mlSubImage.h.

◆ getSourceImageExtent()

ImageVector ml::SubImage::getSourceImageExtent ( ) const

inline

Returns the image extent of the source image that was used to create this SubImage.

This is used for the intersection calculation of the getValidRegion() call.

Definition at line 241 of file mlSubImage.h.

◆ getStride()

ImageVector ml::SubImage::getStride ( ) const

inline

Returns a stride vector to address the memory efficiently.

So the memory address of an subimage voxel can easily be calculated from the dot product of strides and voxel position. Usually stride.x=1, stride.y=x, stride.z=x*y, stride.c=x*y*z, ...

Definition at line 264 of file mlSubImage.h.

◆ getSubImagePointer() [1/2]

void * ml::SubImage::getSubImagePointer ( const ImageVector & voxelPosition ) const

inline

Returns pointer to voxel data of image voxel at 6d voxelPosition relative to the begin of the subimage region.

I.e., the position (0,0,0,0,0,0) is considered the origin of the subimage.

Definition at line 301 of file mlSubImage.h.

◆ getSubImagePointer() [2/2]

void * ml::SubImage::getSubImagePointer	(	MLint	x,
		MLint	y,
		MLint	z ) const

inline

Returns pointer to voxel data of image voxel at 3d position p=(x,y,z) relative to the begin of the subimage region.

I.e., the position (0,0,0,0,0,0) is considered the origin of the subimage.

Definition at line 312 of file mlSubImage.h.

◆ getValidRegion()

SubImageBox ml::SubImage::getValidRegion ( ) const

inline

Returns the valid region of the SubImage, which is defined by the intersection of the SubImage::getBox() and the SubImageBox(ImageVector(0), SubImage::getSourceImageExtent()-1).

The ML host automatically sets the source image extent on SubImages that it creates. If you want to use getValidRegion() on SubImages that you create on your own, you have to call setSourceImageExtent() yourself.

Definition at line 248 of file mlSubImage.h.

References mlrange_cast().

◆ indexToCoord()

static MLEXPORT ImageVector ml::SubImage::indexToCoord	(	MLint	index,
		const ImageVector &	extent )

static

Converts an index into an array with extents extent to a coordinate.

This convenience function is static and can be used also without a class instance. Note that it is not really fast.

◆ isOneValued()

MLEXPORT bool ml::SubImage::isOneValued ( ) const

Returns true if all page voxels have the same value or if the page contains only 1 voxel.

If subimage has no voxels (empty) or more than one value, false is returned.

◆ isValid()

MLEXPORT bool ml::SubImage::isValid ( ) const

Returns 1(=true) if image region and data is valid.

Returns 0(=false) if not. This flag does NOT say anything about the validity of the contents of the memory.

◆ isValidImagePosition() [1/2]

bool ml::SubImage::isValidImagePosition ( const ImageVector & voxelPosition ) const

inline

Returns true if 6d voxelPosition is a valid position within subimage region, i.e., if voxelPosition is within subImgBox.v1 and subImgBox.v2.

If outside, false is returned. So the position p=getBox().v1 addresses the origin of the subimage.

Definition at line 446 of file mlSubImage.h.

◆ isValidImagePosition() [2/2]

bool ml::SubImage::isValidImagePosition	(	MLint	x,
		MLint	y,
		MLint	z ) const

inline

Returns true if 3d position p=(x, y, z) is a valid position within subimage region, i.e., if the position is within subImgBox.v1 and subImgBox.v2.

If outside, false is returned. So the position p=getBox().v1 addresses the origin of the subimage.

Definition at line 458 of file mlSubImage.h.

◆ isValidSubImagePosition() [1/2]

bool ml::SubImage::isValidSubImagePosition ( const ImageVector & voxelPosition ) const

inline

Returns true if 6d voxelPosition is a valid position within subimage region, i.e., if voxelPosition is within (0,0,0,0,0,0) and subimage extents.

If outside, false is returned. So the position p=(0,0,0,0,0,0) addresses the origin of the subimage.

Definition at line 422 of file mlSubImage.h.

◆ isValidSubImagePosition() [2/2]

bool ml::SubImage::isValidSubImagePosition	(	MLint	x,
		MLint	y,
		MLint	z ) const

inline

Returns true if 3d position p=(x, y, z) is a valid position within subimage region, i.e., if the position is within (0,0,0) and subimage extents.

If outside, false is returned. So the position p=(0,0,0,0,0,0) addresses the origin of the subimage and all higher coordinates for c, t and u are always used as 0.

Definition at line 435 of file mlSubImage.h.

◆ operator=()

SubImage & ml::SubImage::operator= ( const SubImage & si )

inline

Assignment operator to get identical copy.

Note that only the pointer to the data is copied, and NOT the data itself.

Definition at line 141 of file mlSubImage.h.

References _box, and mlrange_cast().

◆ setBox() [1/2]

void ml::SubImage::setBox ( const ImageVector & imageExtent )

inline

Sets a rectangular 6d region of the subimage to imageExtent.

So the origin of box will be (0,0,0,0,0,0), the second box edge will result to the extent -1. This overloads the method from base class ImageProperties.

Definition at line 173 of file mlSubImage.h.

References mlrange_cast().

◆ setBox() [2/2]

void ml::SubImage::setBox ( const SubImageBox & subImageBox )

inline

Sets a rectangular 6d region of the subimage to subImageBox.

Overloads method from base class ImageProperties.

Definition at line 163 of file mlSubImage.h.

References mlrange_cast().

◆ setData()

MLEXPORT void ml::SubImage::setData ( void * data )

inline

Sets data as a new memory block for the subimage.

The previous block is NOT touched anymore and it is NOT freed or deleted. If the memory was previously set by setDataFromMemoryBlockHandle(), the memory block handle is cleared.

Definition at line 377 of file mlSubImage.h.

◆ setDataFromMemoryBlockHandle()

MLEXPORT void ml::SubImage::setDataFromMemoryBlockHandle ( const MLMemoryBlockHandle & data )

inline

Sets the managed data from a given MLMemoryBlockHandle, the sub image will store this handle and thus reference count the memory managed by the handle as long as the SubImage stays alive.

This calls setData() with the data from the memory block handle.

Definition at line 383 of file mlSubImage.h.

References MLMemoryBlockHandle::data().

◆ setDataType()

void ml::SubImage::setDataType ( MLDataType dataType )

inline

Overwrite inherited setDataType method to set type of data to dataType.

The validity flag is not changed.

Definition at line 281 of file mlSubImage.h.

References MLSizeOf().

◆ setExtent() [1/2]

void ml::SubImage::setExtent ( const ImageVector & newExtent )

inline

Sets the extent of the subimage (which also changes the strides) to newExtent.

Definition at line 206 of file mlSubImage.h.

References mlrange_cast().

◆ setExtent() [2/2]

void ml::SubImage::setExtent	(	MLint	x,
		MLint	y,
		MLint	z = 1,
		MLint	c = 1,
		MLint	t = 1,
		MLint	u = 1 )

inline

Sets the extent of the subimage (which also changes the strides) to newExtent.

The default value for missing dimensions is 1.

Definition at line 201 of file mlSubImage.h.

◆ setFromImageProperties()

MLEXPORT void ml::SubImage::setFromImageProperties ( const ImageProperties & imageProperties )

Sets the image box and datatype from the imageProperties extent and datatype.

Note that this does not affect the stored data and that the resulting box will have v1 = ImageVector(0) and v2 = imageProperties.getImgExt()-1. This box is adequate for requesting the complete volume defined by imageProperties as a single tile.

◆ setImageExtent()

void ml::SubImage::setImageExtent ( const ImageVector & newExtent )

inline

Deprecated: Use setExtent() instead. Sets the extent of the image.

Definition at line 226 of file mlSubImage.h.

References mlrange_cast().

◆ setOrigin()

void ml::SubImage::setOrigin ( const ImageVector & newOrigin )

inline

Sets the origin of the subimage (which moves the box of the subimage to newOrigin).

This method ensures that the extent stays the same.

Definition at line 190 of file mlSubImage.h.

References mlrange_cast().

◆ setSourceImageExtent()

void ml::SubImage::setSourceImageExtent ( const ImageVector & extent )

inline

Sets the image extent of the source image that was used to create this SubImage to extent.

This is used for the intersection calculation of the getValidRegion() call.

Definition at line 237 of file mlSubImage.h.

◆ toImageProperties()

MLEXPORT ImageProperties ml::SubImage::toImageProperties ( ) const

Converts the Subimage's datatype and extend to an ImageProperties object.

Notice the the min/max values of the ImageProperties are not set, since a Subimage does not store a min/max value.

◆ toStream()

void ml::SubImage::toStream ( std::ostream & ostr ) const

◆ translate()

void ml::SubImage::translate ( const ImageVector & offset )

inline

Translates the box of the subimage by adding the vector offset.

Definition at line 180 of file mlSubImage.h.

Member Data Documentation

◆ _box

SubImageBox ml::SubImage::_box

protected

The box of the subimage.

Definition at line 696 of file mlSubImage.h.

Referenced by operator=().

◆ _data

void* ml::SubImage::_data

protected

Memory chunk managed by this subimage.

Definition at line 702 of file mlSubImage.h.

◆ _dataType

MLDataType ml::SubImage::_dataType

protected

Datatype of the image.

Definition at line 713 of file mlSubImage.h.

◆ _memoryBlock

MLMemoryBlockHandle ml::SubImage::_memoryBlock

protected

Memory block used by this subimage.

Definition at line 705 of file mlSubImage.h.

◆ _sourceImageExtent

ImageVector ml::SubImage::_sourceImageExtent

protected

The extent of the source image, which is used for getValidRegion()

Definition at line 699 of file mlSubImage.h.

◆ _stride

ImageVector ml::SubImage::_stride

protected

Stride vector to address the memory efficiently.

So the memory address of an subimage voxel can easily be calculated from the dot product of strides and voxel position. Usually stride.x=1, stride.y=x, stride.z=x*y, stride.c=x*y*z,...

Definition at line 710 of file mlSubImage.h.

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

MeVis/Foundation/Sources/ML/include/mlSubImage.h

Public Member Functions

Protected Member Functions

Protected Attributes

Properties of subimage

Detailed Description

Constructor & Destructor Documentation

◆ SubImage() [1/4]

◆ SubImage() [2/4]

◆ SubImage() [3/4]

◆ SubImage() [4/4]

◆ ~SubImage()

Member Function Documentation

◆ _calcFillAreaParams()

◆ _copySubImageGeneric()

◆ _getDataTypeInfos()

◆ allocate()

◆ allocateAsMemoryBlockHandle()

◆ calculateMinMax()

◆ compare()

◆ coordToIndex() [1/2]

◆ coordToIndex() [2/2]

◆ copySubImage() [1/2]

◆ copySubImage() [2/2]

◆ fill()

◆ fillBordersWithBorderValues()

◆ fillBordersWithInputValues()

◆ fillBordersWithScalarValue()

◆ fillBordersWithTypeData()

◆ fillInvalidRegionWithBorderValues()

◆ fillInvalidRegionWithScalarValue()

◆ fillInvalidRegionWithTypeData()

◆ fillWithTypeData()

◆ free()

◆ getBox()

◆ getBoxFromExtent()

◆ getBoxFromImageExtent()

◆ getData()

◆ getDataType()

◆ getDataTypeInfos()

◆ getExtent()

◆ getImageExtent()

◆ getImagePointer() [1/2]

◆ getImagePointer() [2/2]

◆ getMemoryBlockHandle()

◆ getNumVoxels()

◆ getOffset()

◆ getOrigin()

◆ getSizeInBytes()

◆ getSourceImageExtent()

◆ getStride()

◆ getSubImagePointer() [1/2]

◆ getSubImagePointer() [2/2]

◆ getValidRegion()

◆ indexToCoord()

◆ isOneValued()

◆ isValid()

◆ isValidImagePosition() [1/2]

◆ isValidImagePosition() [2/2]

◆ isValidSubImagePosition() [1/2]

◆ isValidSubImagePosition() [2/2]

◆ operator=()

◆ setBox() [1/2]

◆ setBox() [2/2]

◆ setData()

◆ setDataFromMemoryBlockHandle()

◆ setDataType()

◆ setExtent() [1/2]

◆ setExtent() [2/2]

◆ setFromImageProperties()

◆ setImageExtent()

◆ setOrigin()

◆ setSourceImageExtent()

◆ toImageProperties()

◆ toStream()

◆ translate()

Member Data Documentation

◆ _box

◆ _data

◆ _dataType

◆ _memoryBlock