【图像融合】用于图像融合方法、客观评估指标、弗里德曼(Friedman)统计检验及其事后检验研究（Matlab代码实现）

 👨‍🎓个人主页：研学社的博客  

💥💥💞💞欢迎来到本博客❤️❤️💥💥

🏆博主优势：🌞🌞🌞博客内容尽量做到思维缜密，逻辑清晰，为了方便读者。

⛳️座右铭：行百里者，半于九十。

📋📋📋本文目录如下：🎁🎁🎁

目录

💥1 概述

📚2 运行结果

2.1 GV_Image_Fusion_Max

2.2  GV_Image_Fusion_PCA

2.3 GV_Image_Fusion_weighted

2.4 basicFusionMethods

2.5 StatisticalComparison

🎉3 参考文献

🌈4 Matlab代码实现

---

​

💥1 概述

本文在不同的多模态图像上实现了一些基本的和最近的融合方法。 使用一些融合指标客观地评估融合方法 弗里德曼统计检验及其事后检验比较了融合方法并检测它们之间的差异。

📚2 运行结果

2.1 GV_Image_Fusion_Max

​

​ 

2.2  GV_Image_Fusion_PCA

​

2.3 GV_Image_Fusion_weighted

​

 ​

2.4和2.5就不一一展示了。 

2.4 basicFusionMethods

2.5 StatisticalComparison

部分代码：

clc
clear
close all
addpath(strcat(pwd,'/functions'));
addpath(strcat(pwd,'/images'));
addpath(strcat(pwd,'/fusion metrics'));

warning('OFF', 'MATLAB:xlswrite:AddSheet');
%% Defining Metrics, Fusion Methods, Source Images, excel file that you want to write the results in
%filter='_allMetrics'; metricNames = {'metric_MI', 'metric_MI2', 'metric_ssim', 'std2', 'entropy', 'metric_Edge_Intensity', 'metric_PSNR', 'metric_AverageGradient',...
%  'metric_Qabf', 'metricXydeas', 'metricCvejic', 'metricPeilla', 'metricChen', 'metricChenBlum', 'metricWang', 'metricYang', 'metricZhao'}' % , 'metricZheng', 'metricPWW', 'metricHossny'}'

filter='_1_12'; metricNames = {'entropy', 'metric_PSNR', 'metric_ssim', 'metric_MI2', 'metric_Qabf', 'metricChen','metricWang', 'metricPeilla', 'metricZhao'}'

choice = questdlg('Which experiments do you like to Evaluate?', ...
    'Menu', ...
    'expr 1-4','expr 5-8', 'expr 1-12', 'expr 1-4');
% Handle response
switch choice
    case 'expr 1-4'
        fusionMethodNames = {'PCNN_NSCT', 'm_PCNN', 'SCM_F', 'NSCT', 'NSCT_SR', 'SCM_M', 'Del_PCA', 'Del_max', 'Del_weighted'}'
        sourceImages = {'G01' 'CT' 'MR_T2' ; 'G02' 'CT' 'MR_T2' ; 'G03' 'MR_T1' 'MR_T2' ; 'G04' 'MR_T1' 'MR_T2'}
        xlsFile = strcat('Qc_1_4',filter,'.xlsx') %write the results to this file
    case 'expr 5-8'
        fusionMethodNames = {'CST','MFDF_NSST','NNSST','ST_NSST', 'FMSAP', 'Del_PCA', 'Del_max', 'Del_weighted'}'
        sourceImages = {'G05' 'CT' 'MR' ; 'G06' 'CT' 'MR' ; 'G07' 'CT' 'MR' ; 'G08' 'CT' 'MR' }
        xlsFile = strcat('Qc_5_8',filter,'.xlsx') %write the results to this file
    case 'expr 1-12'
        %fusionMethodNames = {'FSD', 'GP','DWT','RP', 'MDP', 'PCA', 'LP', 'SIDWT', 'Del_PCA', 'Del_max', 'Del_weighted'}'
        fusionMethodNames = {'FSD', 'GP','DWT','RP', 'MDP', 'LP', 'SIDWT', 'Del_PCA', 'Del_max', 'Del_weighted'}'
        sourceImages = {'G01' 'CT' 'MR_T2' ; 'G02' 'CT' 'MR_T2' ; 'G03' 'MR_T1' 'MR_T2' ; 'G04' 'MR_T1' 'MR_T2' ; ...
            'G05' 'CT' 'MR' ; 'G06' 'CT' 'MR' ; 'G07' 'CT' 'MR' ; 'G08' 'CT' 'MR' ; ...
            'G09' 'CT' 'MR_T2' ; 'G10' 'MR_PD' 'MR_T2' ; 'G11' 'CT' 'MR_GAD' ; 'G12' 'CT' 'MR_T1'}
        xlsFile = strcat('Qc_1_12',filter,'.xlsx') %write the results to this file
        
end
%% initialization
fileName = split(xlsFile,'.xls'); fileName = char(fileName(1));
%%
fn_QC(metricNames,fusionMethodNames,sourceImages, xlsFile);
fn_ANOVA_Friedman(fileName, metricNames,fusionMethodNames,sourceImages)
fn_Friedman_PostHocs(fileName)
fn_metricsCorrelations(fileName, metricNames,fusionMethodNames,sourceImages)
%%
disp('-------------------');
disp(['the Quantitative Comparisons are written in this excel file: ', xlsFile]);
disp(['the ANOVA results are written in this excel file: ', strcat(fileName,'_ANOVA.xlsx')]);
disp(['the Friedman results are written in this excel file: ', strcat(fileName,'_Friedman.xlsx')]);
 

🎉3 参考文献

部分理论来源于网络，如有侵权请联系删除。

[1]李铖.论统一化模型的特设性及缺陷——基于对弗里德曼与科契尔理论的评析[J].自然辩证法研究,2022,38(09):12-18.DOI:10.19484/j.cnki.1000-8934.2022.09.012.

[2]范涛,王昊,李跃艳,邓三鸿.基于多模态融合的非遗图片分类研究[J].数据分析与知识发现,2022,6(Z1):329-337.

🌈4 Matlab代码实现