频域外推EC的matlab代码

频域外推EC的matlab代码，对应的文章是Spatial Error Concealment of Corrupted Image Data using Frequency Selective Extrapolation

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%                                                                  %
%   Copyright （c） 2007 by                                          %
%   Chair of Multimedia Communications and Signal Processing       %
%   University of Erlangen-Nuremberg                               %
%   - all rights reserved -                                        %
%                                                                  %
%   This program is free of charge for scientific and personal     %
%   use. The author （Chair of Multimedia Communications and        %
%   Signal Processing University of Erlangen-Nuremberg） does      %
%   NOT give up his copyright. Any commercial use is prohibited    %
%                                                                  %
%   YOU ARE USING THIS PROGRAM AT YOUR OWN RISK! THE AUTHOR        %
%   IS NOT RESPONSIBLE FOR ANY DAMAGE OR DATA-LOSS CAUSED BY THE   %
%   USE OF THIS PROGRAM.                                           %
%                                                                  %
%                                                                  %
%   If you have any questions please contact:                      %
%                                                                  %
%   Dipl.-Ing. Juergen Seiler                                      %
%   Multimedia Communications and Signal Processing                %
%   University of Erlangen-Nuremberg                               %
%   Cauerstr. 7                                                    %
%   91058 Erlangen Germany                                        %
%                                                                  %
%   email: seiler @ lnt . de                                       %
%                                                                  %
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%


close all;
clear all;


% Simulation parameters
src_img = double（imread（‘lena_y.tif‘））;

border_width = 16;
dEa_threshold = 0;
Ea_threshold = 0;
fft_size = 64;
max_iter = 20;
rho = 0.8;

x_block_size = 16;
y_block_size = 16;

% Setup calculations
y_size = size（src_img1）;
x_size = size（src_img2）;

M = y_block_size+2*border_width;
N = x_block_size+2*border_width;

fft_x_offset = floor（（fft_size-N）/2）;
fft_y_offset = floor（（fft_size-M）/2）;

dE_a= zeros（fft_size）;

% Show original image
figure（1）
imshow（src_img [0 255]）
title（‘Original image‘）


% Generate erroneous image
y_loss_offset = 49:48:y_size-48;
x_loss_offset = 49:48:x_size-48;

loss_img = src_img;
for y_loss_counter = 1:length（y_loss_offset）
	for x_loss_counter = 1:length（x_loss_offset）
		loss_img（y_loss_offset（y_loss_counter）+（1:y_block_size）x_loss_offset（x_loss_counter）+（1:x_block_size）） = 0;
	end
end


% Show erroneous image
figure（2）
imshow（loss_img [0 255]）
title（‘Image with isolated block losses‘）


% Weighting function
w = zeros（fft_size）;
w（fft_y_offset+（1:M） fft_x_offset+（1:N）） =