【图像增强】基于matlab可见边缘梯度比率图像增强【含Matlab源码 1404期】

原创

佛怒火莲 2022-04-06 18:54:27 博主文章分类：matlab完整代码 ©著作权

文章标签 matlab matlab图像处理参考文献图像增强 文章分类 代码人生

©著作权归作者所有：来自51CTO博客作者佛怒火莲的原创作品，请联系作者获取转载授权，否则将追究法律责任

二、部分源代码

%%%% Cleaning
clc
clear all
close all

%%%% Images reading: the input 2 images must be grayscale

NameOri='Original.pgm';
%NameResto='Restored.pgm';
NameResto='Restored2.pgm';
%NameResto='Restored3.pgm';
%NameResto='Restored4.pgm';
%NameResto='Restored5.pgm';


% if the input image is a color image, use following line
% I1=double(rgb2gray(uint8(I1)));



R1=imread(NameResto);
R1=double(R1);
% if the input image is a color image, use following line
% R1=double(rgb2gray(uint8(R1)));

%%%% Figure 1
figure(1)
colormap gray
subplot(1,2,1)
imagesc(I1)
axis image
title('Original')
subplot(1,2,2)
imagesc(R1)
axis image
title('Restored')

%%%% Sobel Gradient
Sy = double([1 2 1;0 0 0;-1 -2 -1]);


Sx = double([1 0 -1; 2 0 -2; 1 0 -1]);
GIx = imfilter(I1,Sx,'symmetric');
GRx1 = imfilter(R1,Sx,'symmetric');

GI=sqrt((GIx.^2)+(GIy.^2));
GR1=sqrt((GRx1.^2)+(GRy1.^2));

minGI=min(GI(:));
maxGI=max(GI(:));

%%%% Figure 2
figure(2)
colormap gray
subplot(1,2,1)
imagesc(GI,[minGI maxGI]);
title(['Gradients of the original image']); 
axis image
colorbar
subplot(1,2,2)
imagesc(GR1,[minGI maxGI]);
title(['Gradients of the restored image']); 
axis image
colorbar

%%%% Contrast Computation at 5%
tic
[C1 Crr1]=functionContrastAt5PerCent(R1);
[Ci Crri]=functionContrastAt5PerCent(I1);
toc

minCrri=min(Crri(:));
maxCrri=max(Crri(:));

%%%% Figure 3
figure(3)
colormap gray
subplot(1,2,1)
imagesc(Crri,[minCrri  maxCrri]);
axis image
title(['Visible edge in the original image']);
colorbar
subplot(1,2,2)
imagesc(Crr1,[minCrri  maxCrri]);
axis image
title(['Visible edge in the restored image']);
colorbar

%%%% Visible Gradients Ratio
Cratio1=zeros(nl,nc);
Cratio1(Crr1>0)=GR1(Crr1>0)./GI(Crr1>0);

rmin=1;
rmax=10;

%%%% Figure 4
figure(4)
imagesc(Cratio1,[rmin rmax]);
axis image
title(['Visible gradients ratio between ',num2str(rmin), ' and ',num2str(rmax)]); 
colormap jet
colorbar

%%%% Descriptor computation

% Rate of new visible edges


% Number of saturated pixels after restoration
ns1=sum(R1(:)==255 |R1(:)==0);
ns1=ns1/(nl*nc);

% Restoration quality (geometric mean ratios of visibility level)
XX=log(Cratio1);
r1=exp((1/(whitePixels1))*nansum(XX(isfinite(XX))));

%%%% Figure 5: Final result with the visible edges at 5% and descriptors
figure(5)
subplot(1,2,1);
colormap gray
imagesc(Ci);
axis image
title(['Visible edges in the original image: ',num2str( whitePixelsi), ' edgels'])
subplot(1,2,2);
colormap gray
imagesc(C1);
axis image
title(['Visible edges in the restored image: ',num2str( whitePixels1), ' edgels'])

axes('position',[0,0,1,1],'visible','off');
text(0.3,0.1,['e=',num2str(e1),'    {\sigma}=',num2str(ns1*100),'%','    r=', num2str(r1),' ' ])