#include <ImFusion/CT/Solver/OptimizationProblem.h>

Optimization problem formulation for tomographic reconstruction. More...

Inheritance diagram for OptimizationProblem:

Detailed Description

Optimization problem formulation for tomographic reconstruction.

Represents minimization problems over image/volume data with data fidelity and regularization terms. Mathematical representation: \(\mbox{argmin}_{x}f(x)\) where \( f(x) = f_0 (x) + \sum_{i = 1}^N w_i f_i (x)\). Here \(f\) is a SharedImageSet valued function with range in the real numbers.

Classes
struct	RegTerm

struct	RegTermOwning

Public Types
using	ExprBase = ImageMath::ExprBase

Public Member Functions
	OptimizationProblem (SharedImageSet &currentSolution, Functional *dataTerm, const std::list< RegTerm > &regTerms={})
	Constructor.

	OptimizationProblem (SharedImageSet &currentSolution, std::unique_ptr< Functional > dataTerm, std::list< RegTermOwning > regTerms={})
	Constructor.

float	eval (const SharedImageSet &in) const
	Evaluates the problem \(f(x) = f_0 (x) + \sum_{i = 1}^N w_i f_i (x)\).

float	eval () const
	Evaluates eval(m_currentSolution)

Status	gradient (const SharedImageSet &in, SharedImageSet &out, const Selection sel=nullptr, const ExprBase exprOut=nullptr) const
	Evaluates the gradient \(\nabla f(in)\).

Status	gradient (SharedImageSet &out, const Selection sel=nullptr, const ExprBase exprOut=nullptr) const
	Evaluates gradient(m_currentSolution, out, sel, exprOut)

template<typename ImplType>
Status	gradient (const SharedImageSet &in, SharedImageSet &out, const Selection sel=nullptr, const ImageMath::ExprBaseTpl< ImplType > exprOut=nullptr) const
	Convenience function that wraps `exprOut` in a sharedPolyWrapper.

template<typename ImplType>
Status	gradient (SharedImageSet &out, const Selection sel=nullptr, const ImageMath::ExprBaseTpl< ImplType > exprOut=nullptr) const
	Convenience function that wraps `exprOut` in a sharedPolyWrapper.

std::unique_ptr< LinearOperator >	hessian (const SharedImageSet &in) const
	Return a LinearOperator representing the hessian matrix \( H_f(in) \).

std::unique_ptr< LinearOperator >	hessian () const
	Evaluates hessian(m_currentSolution)

SharedImageSet &	currentSolution ()
	Getters for m_currentSolutions.

const SharedImageSet &	currentSolution () const

Functional &	dataTerm ()
	Getter for m_dataTerm.

const Functional &	dataTerm () const
	Const-getter for m_dataTerm.

std::list< RegTerm > &	regTerms ()
	Getter for m_regTerms.

const std::list< RegTerm > &	regTerms () const
	Const-getter for m_regTerms.

void	configure (const Properties *p) override
	Configure this object instance by de-serializing the given Properties.

void	configuration (Properties *p) const override
	Serialize the current object configuration into the given Properties object.

Public Member Functions inherited from Configurable
virtual void	configureDefaults ()
	Retrieve the properties of this object, replaces values with their defaults and sets it again.

void	registerParameter (ParameterBase *param)
	Register the given Parameter or SubProperty, so that it will be configured during configure()/configuration().

void	unregisterParameter (const ParameterBase *param)
	Remove the given Parameter or SubProperty from the list of registered parameters.

	Configurable (const Configurable &rhs)

	Configurable (Configurable &&rhs) noexcept

Configurable &	operator= (const Configurable &)

Configurable &	operator= (Configurable &&) noexcept

Protected Attributes
SharedImageSet &	m_currentSolution
	Current solution (non-owning). This may be set by a Solver.

Functional *	m_dataTerm
	Data term of the optimization problem.

std::list< RegTerm >	m_regTerms
	Regularization terms.

std::unique_ptr< Functional >	m_dataTermOwning
	Owning data term of the optimization problem.

std::list< RegTermOwning >	m_regTermsOwning
	Owning regularization terms.

Protected Attributes inherited from Configurable
std::vector< Param >	m_params
	List of all registered Parameter and SubProperty instances.

Additional Inherited Members
Public Attributes inherited from Configurable
Signal	signalParametersChanged
	Emitted whenever one of the registered Parameters' or SubPropertys' `signalValueChanged` signal was emitted.

Constructor & Destructor Documentation

◆ OptimizationProblem() [1/2]

OptimizationProblem	(	SharedImageSet &	currentSolution,
		Functional *	dataTerm,
		const std::list< RegTerm > &	regTerms = {} )

Constructor.

currentSolution must be a value in the domain of dataTerm Domains of dataTerm and of the elements of regTerms must match

◆ OptimizationProblem() [2/2]

OptimizationProblem	(	SharedImageSet &	currentSolution,
		std::unique_ptr< Functional >	dataTerm,
		std::list< RegTermOwning >	regTerms = {} )

Constructor.

currentSolution must be a value in the domain of dataTerm Domains of dataTerm and of the elements of regTerms must match

Member Function Documentation

◆ eval()

float eval ( const SharedImageSet & in ) const

Evaluates the problem \(f(x) = f_0 (x) + \sum_{i = 1}^N w_i f_i (x)\).

Parameters

[in] in SharedImageSet for which the functional \(f\) is evaluated.

Returns: value of the function at the position in.

◆ gradient()

Status gradient	(	const SharedImageSet &	in,
		SharedImageSet &	out,
		const Selection *	sel = nullptr,
		const ExprBase *	exprOut = nullptr ) const

Evaluates the gradient \(\nabla f(in)\).

Parameters

[in]	in	SharedImageSet at which the gradient is evaluated.
[out]	out	gradient.
[in]	sel	Selection on the range of the residual. Can be used for evaluating subsets
[in]	exprOut	If set it will be applied to the result of the gradient. This can e.g. be used to implicitly perform a gradient step with a single function call.

◆ hessian()

std::unique_ptr< LinearOperator > hessian ( const SharedImageSet & in ) const

Return a LinearOperator representing the hessian matrix \( H_f(in) \).

Parameters

[in] in SharedImageSet at which the hessian is evaluated.

◆ configure()

void configure ( const Properties * p )

overridevirtual

Configure this object instance by de-serializing the given Properties.

The default implementation will do so automatically for all registered Parameter and SubProperty instances.

See also: configuration() for the inverse functionality

Reimplemented from Configurable.

◆ configuration()

void configuration ( Properties * p ) const

overridevirtual

Serialize the current object configuration into the given Properties object.

The default implementation will do so automatically for all registered Parameter and SubProperty instances.

See also: configure() for the inverse functionality

Reimplemented from Configurable.

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

ImFusion/CT/Solver/OptimizationProblem.h

Detailed Description

Classes

Public Types

Public Member Functions

Protected Attributes

Additional Inherited Members

Constructor & Destructor Documentation

◆ OptimizationProblem() [1/2]

◆ OptimizationProblem() [2/2]

Member Function Documentation

◆ eval()

◆ gradient()

◆ hessian()

◆ configure()

◆ configuration()