Xmipp
v3.23.11-Nereus
|
Classes | |
class | ARMA_parameters |
Functions | |
void | CausalARMA (MultidimArray< double > &Img, ARMA_parameters &prm) |
void | ARMAFilter (MultidimArray< double > &Img, MultidimArray< double > &Filter, ARMA_parameters &prm) |
void ARMAFilter | ( | MultidimArray< double > & | Img, |
MultidimArray< double > & | Filter, | ||
ARMA_parameters & | prm | ||
) |
ARMAFilter. This function returns the ARMA Filter associated to an ARMA model. This filter should be applied to a random matrix of independent identically distributed variables. In Xmipp the program to do that is Fourierfilter
PARAMETERS: Img - The matrix - Here it's supposed that it comes from an input image Filter - The matrix that will contain the filter. ARParameters - The matrix with the AR model coeficients, as is returned by CausalARMA or NonCausalARMA MAParameters - The matrix with the MA model coeficients, as is returned by CausalARMA or NonCausalARMA dSigma - The Sigma Coeficient for the ARMA model
OUTPUT: The function stores the output in Filter.
DATE: 26-3-2001
Definition at line 194 of file ctf_estimate_psd_with_arma.cpp.
void CausalARMA | ( | MultidimArray< double > & | Img, |
ARMA_parameters & | prm | ||
) |
CausalARMA.
This function determines the coeficients of an 2D - ARMA model
Img(y,x)=sum( AR(p,q)*Img(y+p,x+q)) + sqrt(sigma) * e(y,x)
Where: (p,q) is a vector belonging to a support region N1 AR(p,q) is the matrix with the AR part coeficients of the model sigma is the variance of the random correlated input e(x,y) e(x,y) is random correlated input with zero mean and cross-spectral density:
E[e(x,y)*Img(x+a,y+b)]= sqrt(sigma) if (p,q)=(0,0) sqrt(sigma)*MA(p,q) if (p,q) belongs to N2 0 otherwise
N1 - The support region for the AR part is the first quadrant region defined by N_AR and M_AR N2 - The support region the the MA part is the second quadrant region defined by N_MA and M_MA
This model is determined following the method depicted in:
R. L. Kashyap, "Characterization and Estimation of Two-Dimensional ARMA Models, " IEEE Transactions on Information Theory, Vol. IT-30, No. 5, pp. 736-745, September 1984.
PARAMETERS: Img - The matrix - Here it's supposed that it comes from an image N_AR, M_AR - The order in Rows and Columns directions of the AR part of the model. N_AR, M_AR - The order in Rows and Columns directions of the MA part of the model. ARParameters - The matrix to return the resulting parameteres for the AR part of the model. MAParameters - The matrix to return the resulting parameteres for the MA part of the model.
OUTPUT: The function stores the ARMA parameters into ARParameters and MAParameters Every row of this output matrices has 3 values: 1nd and 2nd- indicate the support point (p,q) for the coeficient 3th column - indicates the value of the coeficient
DATE: 26-3-2001
Definition at line 92 of file ctf_estimate_psd_with_arma.cpp.