Jerome Piovano

Research in Medical Imaging and Computer Vision

segmentation::LocalGaussianRegionModule

LocalGaussianRegionModule Class Template Reference

RegionModule specialization used to drive the LevelSet using a local stats Mumford Shah region evolution. More...

#include <LocalGaussianRegionModule.h>

Inheritance diagram for LocalGaussianRegionModule:

RegionModule SegmentationModule List of all members.

Public Types

typedef RegionModule< DIM,
Pixel, PixelImage > 		base
typedef base::LevelSet LevelSet
typedef base::Index Index
typedef base::ImageType ImageType
typedef BaseImage< DIM, double > ImageFloat
typedef DericheFilter< DIM,
double > 		DericheFilter

Public Member Functions

Constructors - Destructor


 LocalGaussianRegionModule (ImageType *_data, float _weight=1.f, float _sigma=20.0f, bool _var=false)
 Default constructor.
virtual ~LocalGaussianRegionModule ()
 Destructor.
Modify/Access member data


void set_sigma (float _sigma)
 set the variance of the blur
void use_variance (bool _var=true)
 use of local variance
Point dependent 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 measures (const LevelSet &phi, const Index &ind, float meas[])
 miscellaneous measures associated to the module at a point (useful for the assemblage of the modules in the Segmentation)
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_params (const LevelSet &phi)
 update the parameters of the module

Protected Attributes

Member data


float m_sigma
 size of blur
DericheFilter m_deriche
 deriche filter
ImageFloat m_chi_pos
 smoothed binary mask of internal region
ImageFloat m_mean_pos
 local mean of the positive region
ImageFloat m_mean_neg
 local mean of the negative region
ImageFloat m_var_pos
 local variance of the positive region
ImageFloat m_var_neg
 local variance of the negative region
float m_image_variance
 global variance (for speed normalization)
bool m_var
 true if using local variance

Detailed Description

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

RegionModule specialization used to drive the LevelSet using a local stats Mumford Shah region evolution.

Author:
Mikael Rousson
Examples:

2D_LocalGaussian.C.

Definition at line 34 of file LocalGaussianRegionModule.h.

Constructor & Destructor Documentation

LocalGaussianRegionModule ( ImageType *  _data,
float  _weight = 1.f,
float  _sigma = 20.0f,
bool  _var = false 
) [inline]

Default constructor.

Parameters:
_data  image to segment
_weight  global weight of the module
_sigma  variance of gaussian
_var  use of local variance

Definition at line 129 of file LocalGaussianRegionModule.h.

00130                 : RegionModule<DIM, Pixel, PixelImage>(_data, _weight, true), m_sigma(_sigma), m_chi_pos(_data->shape()), m_mean_pos(_data->shape()), m_mean_neg(_data->shape()), m_var_pos(_data->shape()), m_var_neg(_data->shape()), m_image_variance(-1.0f), m_var(_var)
00131 #ifdef SECOND_TERM
00132                 ,m_term1_pos(_data->shape()),m_term2_pos(_data->shape()),m_term1_neg(_data->shape()),m_term2_neg(_data->shape())
00133 #endif
00134 #ifdef BLUR_ODYSSEE_LIB 
00135         ,m_deriche(_sigma, 0)
00136 #else
00137         ,m_deriche(_sigma)
00138 #endif
00139         { }

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

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