SoMultiscaleStructureEnhancementProcessing2d engine More...

#include <ImageViz/Engines/ImageFiltering/TextureFilters/SoMultiscaleStructureEnhancementProcessing3d.h>

Inheritance diagram for SoMultiscaleStructureEnhancementProcessing3d:

Public Types
enum	TensorType { HESSIAN = 0 , GRADIENT = 1 }

enum	Lightness { BRIGHT = 0 , DARK = 1 }

enum	StructureType { ROD = 0 , BALL = 1 , PLANE = 2 }

Public Types inherited from SoImageVizEngine
enum	ComputeMode { MODE_2D = 0 , MODE_3D = 1 , MODE_AUTO = 2 }
	Compute Mode This enum specifies whether the main input will be interpreted as a 3D volume or a stack of 2D images for processing. More...

enum	Neighborhood3d { CONNECTIVITY_6 = 0 , CONNECTIVITY_18 = 1 , CONNECTIVITY_26 = 2 }
	Neighborhood Connectivity 3D. More...

Public Member Functions
	SoMultiscaleStructureEnhancementProcessing3d ()
	Constructor.

Public Member Functions inherited from SoImageVizEngine
virtual SoType	getTypeId () const
	Returns the type identifier for this specific instance.

virtual void	startEvaluate ()
	Evaluate engine and dependencies in another thread without blocking the current one.

virtual void	waitEvaluate ()
	Wait for the end of engine evaluation.

virtual void	abortEvaluate ()
	Abort current processing as soon as possible.

virtual bool	isEvaluating ()
	Returns true if the engine evaluation is in progress.

Public Member Functions inherited from SoEngine
virtual int	getOutputs (SoEngineOutputList &list) const
	Returns a list of outputs in this engine.

SoEngineOutput *	getOutput (const SbName &outputName) const
	Returns a reference to the engine output with the given name.

SbBool	getOutputName (const SoEngineOutput *output, SbName &outputName) const
	Returns (in outputName) the name of the engine output (output).

SoEngine *	copy () const
	Creates and returns an exact copy of the engine.

Public Member Functions inherited from SoFieldContainer
void	setToDefaults ()
	Sets all fields in this object to their default values.

SbBool	hasDefaultValues () const
	Returns TRUE if all of the object's fields have their default values.

SbBool	fieldsAreEqual (const SoFieldContainer *fc) const
	Returns TRUE if this object's fields are exactly equal to fc's fields.

void	copyFieldValues (const SoFieldContainer *fc, SbBool copyConnections=FALSE)
	Copies the contents of fc's fields into this object's fields.

SoNONUNICODE SbBool	set (const char *fieldDataString)
	Sets one or more fields in this object to the values specified in the given string, which should be a string in the Open Inventor file format.

SbBool	set (const SbString &fieldDataString)
	Sets one or more fields in this object to the values specified in the given string, which should be a string in the Open Inventor file format.

void	get (SbString &fieldDataString)
	Returns the values of the fields of this object in the Open Inventor ASCII file format in the given string.

virtual int	getFields (SoFieldList &list) const
	Appends references to all of this object's fields to resultList, and returns the number of fields appended.

virtual int	getAllFields (SoFieldList &list) const
	Returns a list of fields, including the eventIn's and eventOut's.

virtual SoField *	getField (const SbName &fieldName) const
	Returns a the field of this object whose name is fieldName.

virtual SoField *	getEventIn (const SbName &fieldName) const
	Returns a the eventIn with the given name.

virtual SoField *	getEventOut (const SbName &fieldName) const
	Returns the eventOut with the given name.

SbBool	getFieldName (const SoField *field, SbName &fieldName) const
	Returns the name of the given field in the fieldName argument.

SbBool	enableNotify (SbBool flag)
	Notification at this Field Container is enabled (if flag == TRUE) or disabled (if flag == FALSE).

SbBool	isNotifyEnabled () const
	Notification is the process of telling interested objects that this object has changed.

virtual void	setUserData (void *data)
	Sets application data.

void *	getUserData (void) const
	Gets user application data.

Public Member Functions inherited from SoBase
virtual void	touch ()
	Marks an instance as modified, simulating a change to it.

virtual SbName	getName () const
	Returns the name of an instance.

virtual void	setName (const SbName &name)
	Sets the name of an instance.

void	setSynchronizable (const bool b)
	Sets this to be a ScaleViz synchronizable object.

bool	isSynchronizable () const
	Gets the ScaleViz synchronizable state of this object.

Public Member Functions inherited from SoRefCounter
void	ref () const
	Adds a reference to an instance.

void	unref () const
	Removes a reference from an instance.

void	unrefNoDelete () const
	unrefNoDelete() should be called when it is desired to decrement the reference count, but not delete the instance if this brings the reference count to zero.

int	getRefCount () const
	Returns current reference count.

void	lock () const
	lock this instance.

void	unlock () const
	unlock this instance.

Public Member Functions inherited from SoTypedObject
SbBool	isOfType (const SoType &type) const
	Returns TRUE if this object is of the type specified in type or is derived from that type.

template<typename TypedObjectClass >
SbBool	isOfType () const
	Returns TRUE if this object is of the type of class TypedObjectClass or is derived from that class.

Public Attributes
SoSFImageDataAdapter	inImage
	Input image.

SoSFEnum	tensorType
	Defines whether the command will use the gradient tensor or the hessian matrix.

SoSFVec2f	standardDeviationRange
	Standard deviation of the Gaussian kernel at the minimum and maximum scale.

SoSFFloat	standardDeviationStep
	Standard deviation step.

SoSFEnum	lightness
	The lightness type of structures to enhance.

SoSFEnum	structureType
	Shape of structures to extract.

SoImageVizEngineOutput< SoSFImageDataAdapter, SoImageDataAdapter * >	outImage
	Output image.

Public Attributes inherited from SoImageVizEngine
SbEventHandler< EventArg & >	onBegin
	Event raised when the processing begins.

SbEventHandler< EventArg & >	onEnd
	Event raised when processing ends and the result is available.

SbEventHandler< EventArg & >	onProgress
	Event raised while processing is running.

Additional Inherited Members
Static Public Member Functions inherited from SoImageVizEngine
static SoType	getClassTypeId ()
	Returns the type identifier for this class.

Static Public Member Functions inherited from SoEngine
static SoType	getClassTypeId ()
	Returns the type identifier for the SoEngine class.

static SoEngine *	getByName (const SbName &name)
	Looks up engine(s) by name.

static int	getByName (const SbName &name, SoEngineList &list)
	Looks up engine(s) by name.

Static Public Member Functions inherited from SoFieldContainer
static SoType	getClassTypeId ()
	Returns the type of this class.

Static Public Member Functions inherited from SoBase
static SoType	getClassTypeId ()
	Returns type identifier for this class.

Static Public Member Functions inherited from SoTypedObject
static SoType	getClassTypeId ()
	Returns the type identifier for this class.

Detailed Description

SoMultiscaleStructureEnhancementProcessing2d engine

Overview

The purpose of this algorithm is to enhance structures of interest from an image using a multi scale analysis. The result of this algorithm is a score image that can be used with the goal of segmenting the input image.

The SoMultiscaleStructureEnhancement filters compute a score between 0 and 1 for each pixel, 1 representing a good matching with a structure model and 0 a background pixel. This provides a powerful technique for automatically identifying structures such as blood vessels. The score can be computed either on an Hessian matrix to detect ridge structures or on a Gradient tensor for object edges and corners. The structure models available are:

Balls (circular structures with Hessian in 2D, spherical in 3D, object corners with Gradient tensor)
Rods (linear structures)
Plates (only available in 3D)

The principle of the algorithm can be summarized as follows :

A set of tensor fields is extracted from the image by filtering the image at different scales.
A score image is extracted from each tensor field by analyzing the eigenvalues of tensors.
The final score is obtained by selecting the maximum of all score images.

The following publication describes this algorithm when applied to detect Rod structures with the Hessian matrix: A.F.Frangi, W.J.Niessen, K.L.Vincken, M.A.Viergever, "Multiscale vessel enhancement filtering", Lecture Notes in Computer Science(MICCAI), vol. 1496, pp. 130-137, 1998.

Tensor Extraction

2 modes are available for computing the tensor field; the GRADIENT mode and the HESSIAN mode. The first one is based on the gradient tensor the second one on the hessian matrix.

GRADIENT mode

In the GRADIENT mode, the gradient tensor is extracted. This tensor also referred to as structure tensor or second order moment matrix, is a matrix derived from the gradient of the image. This matrix summarizes the predominant directions of the gradient around the voxel of interest.

$S_w(p) = \left( \begin{array}{cc} \int w(r) (I_x(p - r)) ^ 2\, d r & \int w(r) I_x(p - r)I_y(p - r)\, d r \\ \int w(r) I_x(p - r)I_y(p - r)\, d r & \int w(r) (I_y(p - r)) ^ 2\, d r \end{array} \right)$

Where $w(r)$ is a gaussian kernel controlling the scale of analysis.

HESSIAN mode

In this mode, the tensor field is based on the extraction of the hessian matrix of the image. This hessian matrix is computed by filtering the image with the derivative of a gaussian kernel.The standard deviation of the kernel controls the scale of analysis.

Feature extraction

The parameter structureType controls the type of structure that is extracted from the tensor field(ROD/BALL/PLANE). The computation is based on eigen values $\lambda_{1}$ , $\lambda_{2}$ and $\lambda_{3}$ where $\mid \lambda_{1} \mid \geq \mid \lambda_{2} \mid \geq \mid \lambda_{3} \mid$ .

Let first introduce

$\mathcal{R}_b = \frac{ \lambda_1 }{ \sqrt{ \lambda_2 \lambda_3 } }$ , and

$\mathcal{R}_a = \frac{ \lambda_2 }{ \lambda_3 }$

The first ratio attains its maximum for blob-like structure. The second ratio distinguishes plate-like and line-like structures.

The score for StructureType3D::ROD corresponds to :

$\text{score}_{\text{ROD}} = (1-\exp(-\frac{\mathcal{R}_a^2}{2\alpha^2}))\exp(-\frac{\mathcal{R}_b^2}{2\beta^2})(1-\exp(-\frac{S^2}{2c^2}))$

The score for StructureType3D::BALL is computed as follow :

$\text{score}_{\text{BALL}} = (1-\exp(-\frac{\mathcal{R}_a^2}{2\alpha^2}))(1-\exp(-\frac{\mathcal{R}_b^2}{2\beta^2}))(1-\exp(-\frac{S^2}{2c^2}))$

The score for StructureType3D::PLANE is computed as follow :

$\text{score}_{\text{PLANE}} = \exp(-\frac{\mathcal{R}_a^2}{2\alpha^2})\exp(-\frac{\mathcal{R}_b^2}{2\beta^2})(1-\exp(-\frac{S^2}{2c^2}))$

where $S$ is the tensor norm $S = \sqrt{\sum \lambda_i^2}$ , $\alpha$ is a threshold which controls the flatness sensitivity, $\beta$ is a threshold which controls the blobness sensitivity, and $ c $ is a sensitivity threshold which controls the noise influence. $c = \text{noiseCutoff} \times S_{max}$ with $S_{max}$ the maximum Hessian norm in the image.

In the HESSIAN mode, the lightness parameter limits the feature extraction to dark or bright objects by analyzing the sign of eigenvalues. This parameter is ignored in GRADIENT mode where tensors are positive definite matrices.