GeoCuts Markov Random Fields

Segmentation by minimum surfaces (object label = 2, background label = 3) More...

Detailed Description

Segmentation by minimum surfaces (object label = 2, background label = 3)

Warning

some annotated functions are tests functions, no guarantee on the results !!!

Author

Jean Stawiaski

Collaboration diagram for GeoCuts Markov Random Fields:

Functions
template<class T1 , class T2 >
RES_T	geoCutsMRF_Ising (const Image< T1 > &imIn, const Image< T2 > &imMarker, double Beta, double Sigma, const StrElt &nl, Image< T2 > &imOut)
	Markov Random Fields segmentation with two labels (2 labels, object = 2, background = 3) with The Ising Model. More...
template<class T1 , class T2 >
RES_T	geoCutsMRF_EdgePreserving (const Image< T1 > &imIn, const Image< T2 > &imMarker, double Beta, double Sigma, const StrElt &nl, Image< T2 > &imOut)
	Markov Random Fields segmentation with two labels (2 labels, object = 2, background = 3) with edge preserving prior. More...
template<class T1 , class T2 >
RES_T	geoCutsMRF_Potts (const Image< T1 > &imIn, const Image< T2 > &imMarker, double Beta, double Sigma, const StrElt &nl, Image< T2 > &imOut)
	Multi-Label MAP Markov Random Field with Ising prior (Potts Model) More...

Function Documentation

geoCutsMRF_Ising()

RES_T geoCutsMRF_Ising	(	const Image< T1 > &	imIn,
		const Image< T2 > &	imMarker,
		double	Beta,
		double	Sigma,
		const StrElt &	nl,
		Image< T2 > &	imOut
	)

Markov Random Fields segmentation with two labels (2 labels, object = 2, background = 3) with The Ising Model.

Note

: For Beta and sigma parameters, read Markov Random Fields section on jean Stawiaski thesis

Parameters

[in]	imIn	: Image<T> in
[in]	imMarker	: Image<T> marker
[in]	Beta
[in]	Sigma
[in]	nl	: Neighborlist
[out]	imOut	: Image<T> out

geoCutsMRF_EdgePreserving()

RES_T geoCutsMRF_EdgePreserving	(	const Image< T1 > &	imIn,
		const Image< T2 > &	imMarker,
		double	Beta,
		double	Sigma,
		const StrElt &	nl,
		Image< T2 > &	imOut
	)

Markov Random Fields segmentation with two labels (2 labels, object = 2, background = 3) with edge preserving prior.

Note

: For Beta and sigma parameters, read Markov Random Fields section on jean Stawiaski thesis

Parameters

[in]	imIn	: Image<T> in
[in]	imMarker	: Image<T> marker
[in]	Beta
[in]	Sigma
[in]	nl	: Neighborlist
[out]	imOut	: Image<T> out

geoCutsMRF_Potts()

RES_T geoCutsMRF_Potts	(	const Image< T1 > &	imIn,
		const Image< T2 > &	imMarker,
		double	Beta,
		double	Sigma,
		const StrElt &	nl,
		Image< T2 > &	imOut
	)

Multi-Label MAP Markov Random Field with Ising prior (Potts Model)

Note

: For Beta and sigma parameters, read Markov Random Fields section on jean Stawiaski thesis

Parameters

[in]	imIn	: Image<T> in
[in]	imMarker	: Image<T> marker
[in]	Beta
[in]	Sigma
[in]	nl	: Neighborlist
[out]	imOut	: Image<T> out