Main Page | Namespace List | Class Hierarchy | Alphabetical List | Class List | Directories | File List | Namespace Members | Class Members | File Members

JointHistogram Class Reference

Computing joint histograms, and associated intensity normalization. More...

#include <Normalization.hpp>

Collaboration diagram for JointHistogram:

Collaboration graph
[legend]
List of all members.

Public Member Functions

name conversion from bin number
to real value float 
SrcBinF (float v) const
 index of src bin for src value=v
float RefBinF (float v) const
int SrcBin (float v) const
int RefBin (float v) const
double SrcVal (int x) const
double SrcVal (double x) const
double RefVal (int y) const
double RefVal (double y) const
int SSrcBin (float v) const
int SRefBin (float v) const
float SrcLimit (float v) const
float RefLimit (float v) const
void AddDisplayGrid ()
 Adds a display grid to joint histogram image. Only for visualisation purposes.
float GetSrcBinTotalProba (int x) const
 Computes total probability of a source bin (adds all corresponding ref bins).
Signal1Df MedianTransferFct () const
 Compute median normalization intensity transfer fct (DEPRECATED!!!).
Signal1Df MaximumProbabilityTransferFct () const
 Compute maximum probability intensity transfer fct.
Signal1Df BestFitTransferFct (double smoothingCoef=1, bool allowDecreasing=false) const
 Compute intensity transfer fct using gsl minimisation, this is intended to replace the other methods.
Signal1Df ConfidenceWeights () const
 Computes confidence weights for each source intensity.
 JointHistogram (const string &filename)
 JointHistogram (const Image3Df &src, const Image3Df &ref, int _nbSrcBins=100, int _nbRefBins=400)

Static Public Member Functions

static Signal1Df SmoothAndExtend (const Signal1Df &src, const Signal1Df &weights, double smoothSigma, double outlierSigma)
 smooths a signal using confidence weights

Public Attributes

int nbSrcBins
int nbRefBins
Image3Df histogramImage
double srcMin
 min value for src image
double srcMax
 max value for src image
double refMin
 min value for ref image
double refMax
 max value for ref image

Protected Member Functions

void ComputeFromIndependentSamples (const Image3Df &src, const Image3Df &ref)
 Simple computation of joint histo.
void ComputeFromDirectionalLinearInterpolation (const Image3Df &src, const Image3Df &ref, int dir)
 Computation of joint histo using *linear* interpolation (smoother, much better quality results).
virtual void FindRanges (const Image3Df &src, const Image3Df &ref)
 Determine ranges (srcMin,srcMax) and (refMin,refMax).
 JointHistogram (int _nbSrcBins=100, int _nbRefBins=400)
 Ctor for derived classes.

Detailed Description

Computing joint histograms, and associated intensity normalization.

Definition at line 57 of file Normalization.hpp.


Member Function Documentation

Signal1Df JointHistogram::BestFitTransferFct double  smoothingCoef = 1,
bool  allowDecreasing = false
const
 

Compute intensity transfer fct using gsl minimisation, this is intended to replace the other methods.

it should be better...

Definition at line 712 of file Normalization.cpp.

References Signal1Df::Size().

Signal1Df JointHistogram::SmoothAndExtend const Signal1Df src,
const Signal1Df weights,
double  smoothSigma,
double  outlierSigma
[static]
 

smooths a signal using confidence weights

This is a complex algorithm: it predicts values from tangents at each point, smooths the predicted values, and the repeats the operation with outlier rejection, and less smoothing

Definition at line 36 of file Normalization.cpp.

References Signal1Df::Size(), and Signal1Df::Smooth().


The documentation for this class was generated from the following files:
Generated on Fri Jun 17 13:36:19 2005 for ImLib3D by  doxygen 1.4.2