Décomposition en curvelet

[invite805bfddf]

salut,

J'essaye d'appliquer la décomposition en curvelet pour l'image de huffman,

Pour cela j'applique un bruit à l'image originale puis je calcul le psnr entre l'image originale et celle débruité.

Je vous donne le code matlab que j'ai utilisé :

close all; clear all; 
clc

im = double(imread('C:\Users\user\Desktop\ACTIVITEMULTI\base_originale\Hoffman.jpg'));
%s = 512;

sigma = 20; % NV(in);
noise = randn(size(im));
noise = sigma.*(noise./(std(noise(:))));

% cfg =  [4 5 5 6];
%cfg =  [3 4 4 5];
cfg =  [2 2 2 2 2];
alpha = 0.15;
s = 512;
resi = false;

%%%%%%%%%%PDTDFB_WIN   Generate the Freq. windows that used in PDTDFB toolbox%%
%
%	F2 = pdtdfb_win(s2, alpha, Lnlev, res)
%
% Input:
%   s : size of the generated window
%   alpha : paramter for meyer window
%
% Output:
%   FL: 2-D window of low pass function for lower resolution curvelet
%   F : cell of size n containing 2-D matrices of size s*s
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

F = pdtdfb_win(s, alpha, length(cfg), resi);

% 
disp('Compute all thresholds');
Ff = ones(s);
X = fftshift(ifft2(Ff)) * sqrt(prod(size(Ff)));
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% PDTDFBDEC_F   Pyramidal Dual Tree Directional Filter Bank Decomposition
% using FFT at the multresolution FB and the first two level of dual DFB
% tree
%
%	y2 = pdtdfbdec_f(im, nlevs, F2, alpha, res)
%
% Input:
%   x:      Input image
%   nlevs:  vector of numbers of directional filter bank decomposition levels
%           at each pyramidal level (from coarse to fine scale).
%           0 : Laplacian pyramid level
%           1 : Wavelet decomposition
%           n : 2^n directional PDTDFB decomposition
%   F2:     [optional] Precomputed F window 
%   alpha:  [optional] Parameter for F2 window, incase F2 is not pre-computed
%   res :   [optional] Boolean value, specify weather existing residual
%           band
%
% Output:
%   y:      a cell vector of length length(nlevs) + 1, where except y{1} is
%           the lowpass subband, each cell corresponds to one pyramidal
%           level and is a cell vector that contains bandpass directional
%           subbands from the DFB at that level.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
yn = pdtdfbdec_f(noise, cfg, F, alpha, resi);
y = pdtdfbdec_f(im, cfg, F, alpha, resi);
clear eng


for scale = 1:length(cfg)
    for dir = 1:2^cfg(scale)
        Y_coef_real = y{scale+1}{1}{dir};
        % imaginary part
        Y_coef_imag = y{scale+1}{2}{dir};
        % Signal variance estimation
        Y_coef = Y_coef_real+j*Y_coef_imag;
        
        Y_noise_real = yn{scale+1}{1}{dir};
        % imaginary part
        Y_noise_imag = yn{scale+1}{2}{dir};
        % Signal variance estimation
        Y_noise = Y_noise_real+j*Y_noise_imag;
        
        Y_coef = abs(Y_noise).*Y_coef./abs(Y_coef);

        y{scale+1}{1}{dir} = real(Y_coef);
        y{scale+1}{2}{dir} = imag(Y_coef);

    end
end
y{1} = zeros(size(y{1}));

% Inverse Transform
tic
pim = pdtdfbrec_f(y, F, alpha,resi);toc
respdtdfb = psnr(im, pim)

quand je l'exécute j'ai le message d'erreur suivant


?? Error using ==> times
Matrix dimensions must agree.

Error in ==> pdtdfbdec_f at 74
imf2 = imf.*F{1};

Error in ==> curvelette at 62
yn = pdtdfbdec_f(noise, cfg, F, alpha, resi);

Je serais reconnaissante à vous si vous pouvez m'aider.

Merci d'avance.
-----

[inviteb9f49292]

Le message d'erreur est pourtant explicite, tu essaies de multiplier deux matrices entre elles dont les dimensions ne correspondent pas i.e. le nb de colonne de l'une n'est pas égale au nb de lignes de l'autre et/ou vice-versa

[invite805bfddf]

oui j'ai compris le message d'erreur mais j'ai pas pu corriger leurs dimension.

Comment faire??