DeblurAlgorithm3.m

function [xEst, wEst] = DeblurAlgorithm3(w,x,y,kernel_size)
%UNTITLED5 Summary of this function goes here
%   Detailed explanation goes here
addpath(genpath('./Graph_Based_BID_p1.1'));

final = uint8(255.0 * mat2gray(y));
imwrite(final,'temp.png');
I=imread('./temp.png');

if numel(size(I))>2
    Y_b=im2double(rgb2gray(I));
else
    Y_b=im2double(I);
end
figure(2);
imshow(Y_b);
title('Blurry Image');
% Blind image deblurring
k_estimate_size=kernel_size; % set approximate kernel size
show_intermediate=true; % show intermediate output or not

border=1;% cut boundaries of the input blurry image (default border length: 20)
tic;
[ k_estimate,Y_intermediate ] = bid_rgtv_c2f_cg( Y_b(border+1:end-border,border+1:end-border),k_estimate_size,show_intermediate );
t=toc;
fprintf('Grpah based Blind Deblurring Running Time:%f s\n',t);

% Fast Hyper Laplacian Priors non-blind deblurring
I_b=im2double(I);
I_FHLP=I_b;
I_blur=I_b;
tic;
if numel(size(I_FHLP))>2
    for i=1:3
        [ I_FHLP(:,:,i) ]=Deconvolution_FHLP(I_blur(:,:,i),k_estimate);
    end
else
    [ I_FHLP ]=Deconvolution_FHLP(I_blur,k_estimate);
end
t=toc;
fprintf('FHLP Run Time: %f s\n',t);
xEst = I_FHLP;
wEst = k_rescale(k_estimate);
end