zxhTransformFFD Class Reference

transformation type: Free form deformation file extension name:.FFD, currently save as text file, will change to imagedata files in the future zxhtodo todo More...

#include <zxhTransformFFD.h>

Inheritance diagram for zxhTransformFFD:

List of all members.

Public Member Functions
	zxhTransformFFD (void)
	constructor
virtual std::string	GetTransformType ()
virtual std::string	GetPrintString ()
	base-> size[i] each line->spacing[i] each line->offsets_of_controlpoints[t][z][y] each line
virtual zxhTransformBase *	Clone (zxhTransformBase *&pRet)
virtual bool	ComputeMetricGradient (zxhMetricBase *pMetric, zxhTransformBase *pInfluence=0)
virtual bool	ComputeMetricGradientOnDirection (zxhMetricBase *pMetric, zxhTransformBase *pInfluence, zxhImageModelingBaseT< float, float > *pDirectionFFDsNVModelRef)
virtual bool	PrePostComputeMetricGradient (zxhMetricBase *pMetric, zxhTransformBase *pInfluence, bool isPre)

---

Detailed Description

transformation type: Free form deformation file extension name:.FFD, currently save as text file, will change to imagedata files in the future zxhtodo todo

---

Member Function Documentation

zxhTransformBase * zxhTransformFFD::Clone

(

zxhTransformBase *&

pRet

)

[virtual]

Returns:

Reimplemented from zxhTransformFFDBase.

bool zxhTransformFFD::ComputeMetricGradient	(	zxhMetricBase *	pMetric,
		zxhTransformBase *	pInfluence = 0
	)		[virtual]

methods for gradient type objects computer gradient, influence control status for optimisation if infl==0, then use m_fDifferentialStep if pDifferentialStep==0,then using m_fDifferentialStep

Implements zxhGradientBase.

bool zxhTransformFFD::ComputeMetricGradientOnDirection	(	zxhMetricBase *	pMetric,
		zxhTransformBase *	pInfluence,
		zxhImageModelingBaseT< float, float > *	pDirectionFFDsNVModelRef
	)		[virtual]

the control point only move along the normal vector of the surface image in the source image space 1) every step optimisation, the FFDs needs to be identity ComputeMetricGradientOnDirection, the NVc is computed from original NVo Using $NV_c = Normalise( NV_o Jacobian(m_pConcatenatedTransformsByRegridding) )$ 2) after each optimisation step, in Optimizer::AdjustParameters() regridding concatenate newly optimised FFDs to m_pConcatenatedTransformsByRegridding

virtual std::string zxhTransformFFD::GetTransformType

(

)

[inline, virtual]

Returns:

Reimplemented from zxhTransformFFDBase.

bool zxhTransformFFD::PrePostComputeMetricGradient	(	zxhMetricBase *	pMetric,
		zxhTransformBase *	pInfluence,
		bool	isPre
	)		[virtual]

adjust influence and differential virtual bool ComputeAdjustment( zxhTransformFFDBase * pInfl ) {return false; } ;

Reimplemented from zxhGradientFFDBase.

---

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

• zxhTransformFFD.h
• zxhTransformFFD.cpp