[ VIGRA Homepage | Function Index | Class Index | Namespaces | File List | Main Page ]
Image Correlation |
Functions | |
template<... > | |
void | crossCorrelation (...) |
This function performes a (slow) cross-correlation. More... | |
template<... > | |
void | fastCrossCorrelation (...) |
This function performes a fast cross-correlation. More... | |
template<... > | |
void | fastNormalizedCrossCorrelation (...) |
This function performes a fast normalized cross-correlation. More... | |
template<... > | |
void | normalizedCrossCorrelation (...) |
This function performes a (slow) normalized cross-correlation. More... | |
Fast (FFT-based) and slow algorithms to estimate the correlation between images.
void vigra::fastCrossCorrelation | ( | ... | ) |
This function performes a fast cross-correlation.
This function performes a fast cross-correlation using the Fast Fourier Transform and the dependency of the convolution and the correlation in Fourier space.
The input pixel type T1
must be a linear space over the window functions' value_type T
, i.e. addition of source values, multiplication with functions' values, and NumericTraits must be defined. The mask's value_type must be an algebraic field, i.e. the arithmetic operations (+, -, *, /) and NumericTraits must be defined.
By default, the borders are filled with zeros. Use the clearBorders switch to change that behavior if you need to.
Declarations:
pass 2D array views:
Usage:
#include <vigra/correlation.hxx>
Namespace: vigra
Preconditions:
The image must be larger than the size of the mask.
void vigra::fastNormalizedCrossCorrelation | ( | ... | ) |
This function performes a fast normalized cross-correlation.
To compute the normalized cross correlation in a fast way, it is using the Fast Fourier Transform and sum-image look-up-tables as it is suggested by J.P.Lewis (1995): "Fast Normalized Cross-Correlation".
The input pixel type T1
must be a linear space over the window functions' value_type T
, i.e. addition of source values, multiplication with functions' values, and NumericTraits must be defined. The mask's value_type must be an algebraic field, i.e. the arithmetic operations (+, -, *, /) and NumericTraits must be defined.
By default, the borders are filled with zeros. Use the clearBorders switch to change that behavior if you need to.
Declarations:
pass 2D array views:
Usage:
#include <vigra/correlation.hxx>
Namespace: vigra
Preconditions:
The image must be larger than the size of the mask.
void vigra::crossCorrelation | ( | ... | ) |
This function performes a (slow) cross-correlation.
This function performes a (slow) cross-correlation using the window function environment and comparison of the mask with the underlying image part for each pixel. This may however be faster for very few comparisons.
The input pixel type T1
must be a linear space over the window functions' value_type T
, i.e. addition of source values, multiplication with functions' values, and NumericTraits must be defined. The mask's value_type must be an algebraic field, i.e. the arithmetic operations (+, -, *, /) and NumericTraits must be defined.
Declarations:
pass 2D array views:
Usage:
#include <vigra/correlation.hxx>
Namespace: vigra
Preconditions:
The image must be larger than the size of the mask.
void vigra::normalizedCrossCorrelation | ( | ... | ) |
This function performes a (slow) normalized cross-correlation.
This function performes a (slow) normalized cross-correlation using the window function environment and comparison of the mask with the underlying image part for each pixel. This may however be faster for very few comparisons.
The input pixel type T1
must be a linear space over the window functions' value_type T
, i.e. addition of source values, multiplication with functions' values, and NumericTraits must be defined. The mask's value_type must be an algebraic field, i.e. the arithmetic operations (+, -, *, /) and NumericTraits must be defined.
Declarations:
pass 2D array views:
Usage:
#include <vigra/correlation.hxx>
Namespace: vigra
Preconditions:
The image must be larger than the size of the mask.
© Ullrich Köthe (ullrich.koethe@iwr.uni-heidelberg.de) |
html generated using doxygen and Python
|