Save Extracted Features (For Loop)

12 次查看（过去 30 天）
Gigi N
Gigi N 2021-7-31
回答： Image Analyst 2021-7-31
HI, I have the below code to perform a DWT and then perform feature extractions. However, I'm having difficulty saving the extraction values into .xlsx format because every new entry gets overwritten, and the 1st column values and 2nd column 1st row values are wrong (scroll down to the last part for the xlsx result). Please advise how I can go about doing this as I'm very new to MATLAB. Thank you.
srcFile = dir('filepath');
for i = 1:length(srcFile)
filename = strcat('filepath',srcFile(i).name);
j = imread(filename);
[xar,xhr,xvr,xdr] = dwt2(j(:,:,1),'bior5.5');
[xag,xhg,xvg,xdg] = dwt2(j(:,:,1),'bior5.5');
[xab,xhb,xvb,xdb] = dwt2(j(:,:,1),'bior5.5');
xa(:,:,1) = xar ; xa(:,:,2) = xag ; xa(:,:,3) = xab;
xh(:,:,1) = xhr ; xh(:,:,2) = xhg ; xh(:,:,3) = xhb;
xv(:,:,1) = xvr ; xv(:,:,2) = xvg ; xv(:,:,3) = xvb;
xd(:,:,1) = xdr ; xd(:,:,2) = xdg ; xd(:,:,3) = xdb;
[xaar,xhhr,xvvr,xddr] = dwt2(xa(:,:,1),'bior5.5');
[xaag,xhhg,xvvg,xddg] = dwt2(xa(:,:,1),'bior5.5');
[xaab,xhhb,xvvb,xddb] = dwt2(xa(:,:,1),'bior5.5');
xaa(:,:,1) = xaar ; xaa(:,:,2) = xaag ; xaa(:,:,3) = xaab;
xhh(:,:,1) = xhhr ; xhh(:,:,2) = xhhg ; xhh(:,:,3) = xhhb;
xvv(:,:,1) = xvvr ; xvv(:,:,2) = xvvg ; xvv(:,:,3) = xvvb;
xdd(:,:,1) = xddr ; xdd(:,:,2) = xddg ; xdd(:,:,3) = xddb;
DWT_feat = [xaa,xhh,xvv,xdd];
gray = rgb2gray(DWT_feat);
glcm = graycomatrix(gray);
stats = graycoprops(glcm,'Contrast Correlation Energy Homogeneity');
Contrast = stats.Contrast;
Correlation = stats.Correlation;
Energy = stats.Energy;
Homogeneity = stats.Homogeneity;
Mean = mean2(DWT_feat);
StdDev = std2(DWT_feat);
Entropy = entropy(DWT_feat);
RMS = mean2(rms(DWT_feat));
Variance = mean2(var(double(DWT_feat)));
a = sum(double(DWT_feat(:)));
Smoothness = 1-(1/(1+a));
Kurtosis = kurtosis(double(DWT_feat(:)));
Skewness = skewness(double(DWT_feat(:)));
VariableNames = {'Contrast','Correlation', 'Energy', 'Homogeneity', 'Mean', 'StdDev', 'Entropy', 'RMS', 'Variance', 'Smoothness', 'Kurtosis', 'Skewness'};
T = table(Contrast(:), Correlation(:), Energy(:), Homogeneity(:), Mean(:), StdDev(:), Entropy(:), RMS(:), Variance(:), Smoothness(:), Kurtosis(:), Skewness(:));
AT = array2table(T,'VariableNames', {'Contrast','Correlation', 'Energy', 'Homogeneity', 'Mean', 'StdDev', 'Entropy', 'RMS', 'Variance', 'Smoothness', 'Kurtosis', 'Skewness'});
disp(AT);
xlswrite('savepath',AT,'Sheet1','A1');
end
Result (which is wrong):
A T Energy Homogeneity Mean StdDev Entropy RMS Variance Smoothness Kurtosis Skewness
T 0.720361 0.3986 0.871775 24.72161 95.91721 1.395854 37.76407 4277.009 1 32.69805 5.19293

请先登录，再回答此问题。

回答（1 个）

Image Analyst
Image Analyst 2021-7-31
You need to index the variables that you want to save a value for on every iteration. For example:
Entropy(i) = entropy(DWT_feat);
but it's better not to use i (the imaginary variable) as a loop counter, Use k instead.
Also, don't use built-in function names, like gray(), as a variable name. Use grayImage instead.
Next put in comments before you write the code, then write the code. Or at least put in comments AS you write the code. But to not have any comments at all would not fly in my company.
Finally, when you post code, you can highlight it and click the code icon. That will format it properly and allow us to copy the whole code with a single click so we can try it out on our computers.

类别

在 Help Center 和 File Exchange 中查找有关 Resampling Techniques 的更多信息

标签

产品


版本

R2019a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by