Parameters for image inv_model.gaussian_prior_likelihood.img_mean -> image mean inv_model.gaussian_prior_likelihood.R_prior -> L*L' = inv(image covariance) inv_model.gaussian_prior_likelihood.img_exp -> ( default = 2) Parameters for data inv_model.gaussian_prior_likelihood.Noise -> L*L' = inv(Noise covariance) inv_model.gaussian_prior_likelihood.data_exp -> ( default = 2)
0001 function likelihood= gaussian_prior_likelyhood( inv_model, x, y, J ) 0002 % Parameters for image 0003 % inv_model.gaussian_prior_likelihood.img_mean -> image mean 0004 % inv_model.gaussian_prior_likelihood.R_prior -> L*L' = inv(image covariance) 0005 % inv_model.gaussian_prior_likelihood.img_exp -> ( default = 2) 0006 % Parameters for data 0007 % inv_model.gaussian_prior_likelihood.Noise -> L*L' = inv(Noise covariance) 0008 % inv_model.gaussian_prior_likelihood.data_exp -> ( default = 2) 0009 0010 0011 x_m = inv_model.gaussian_prior_likelihood.img_mean; 0012 L_x = inv_model.gaussian_prior_likelihood.R_prior; 0013 p_x = inv_model.gaussian_prior_likelihood.img_exp; 0014 L_n = inv_model.gaussian_prior_likelihood.Noise; 0015 p_n = inv_model.gaussian_prior_likelihood.data_exp; 0016 0017 img_residual= x - x_m; 0018 data_residual= y - J*x; 0019 0020 likelihood= exp(- norm(L_n * data_residual, p_n) ... 0021 - norm(L_x * img_residual, p_x));