Jerome Piovano

Research in Medical Imaging and Computer Vision

• Home
• Research
• Publications
• Teaching
• Biography
• Shared code and data

GaussianRegionModule Class Template Reference

RegionModule specialization used to drive the LevelSet using a scalar gaussian region evolution. More...

#include <GaussianRegionModule.h>

Inheritance diagram for GaussianRegionModule:

List of all members.

Public Types
typedef RegionModule< DIM, Pixel, PixelImage >	base
typedef base::LevelSet	LevelSet
typedef base::Index	Index
typedef base::Histogram	Histogram
typedef base::ImageType	ImageType
Public Member Functions
Constructors - Destructor
	GaussianRegionModule (ImageType *_data, float _weight=1.f, bool _var=false, bool _adaptive=true)
	Default constructor.
virtual	~GaussianRegionModule ()
	Destructor.
Modify/Access member data
void	use_variance (bool _var=true)
	false => Chan-Vese piecewise constant mumford shah evolution
Member functions
float	value (const LevelSet &phi, const Index &ind)
	Value of the module at a point.
float	energy (const LevelSet &phi, const Index &ind)
	Energy of the module at a point.
void	init (const LevelSet &phi)
	Initialize the module.
void	update (const LevelSet &phi)
	update the module
void	save_result (const LevelSet &phi, std::string evodir, int iter)
	save results depending on the module data
void	update_gaussian_params ()
	update the parameters of the gaussian (mean, variance) using the histograms, and precompute the log-likelihood of all possible values
Protected Attributes
Member data
std::vector< Pixel >	m_ll_pos
	Log-likelihoods for the positive region.
std::vector< Pixel >	m_ll_neg
	Log-likelihoods for the negative region.
float	m_mean_pos
	mean of the positive region
float	m_mean_neg
	mean of the negative region
float	m_var_pos
	variance of the positive region
float	m_var_neg
	variance of the negative region
bool	m_var
	Use of the variance.

---

Detailed Description

template<unsigned DIM, typename Pixel = float, typename PixelImage = float>
class segmentation::GaussianRegionModule< DIM, Pixel, PixelImage >

RegionModule specialization used to drive the LevelSet using a scalar gaussian region evolution.

Author:

Mikael Rousson

Examples:

2D_Gaussian.C, 2D_LocalGaussian.C, 2D_SoftShapePrior.C, and 3D_ChanVese.C.

Definition at line 25 of file GaussianRegionModule.h.

---

Constructor & Destructor Documentation

GaussianRegionModule	(	ImageType *	_data,
		float	_weight = 1.f,
		bool	_var = false,
		bool	_adaptive = true
	)			[inline]

Default constructor.

Parameters:

_data	image to segment
_weight	weight of the module
_var	false => don't use the variance (Chan-Vese evolution)
_adaptive	if the module is adaptive

Definition at line 102 of file GaussianRegionModule.h.

                :RegionModule<DIM, Pixel, PixelImage>(_data, _weight, _adaptive), m_mean_pos(0), m_mean_neg(0), m_var_pos(0), m_var_neg(0), m_var(_var)
        { }

---

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

• /home/jpiovano/src/Odyssee++/trunk/Libs/LevelSet/src/SegmentationModule/GaussianRegionModule.h

For further information, please contact Jerome Piovano - Last update 2008-02-08