extracting HOG features from .mat files and usage of "fitcecoc"

5 次查看(过去 30 天)
I have '.mat files' for different objects and I want to extract HOG features from the mat files, and I want to apply those features on "fitcecoc" SVM one vs one classifier. I have written a code but giving the error like this:
CS=[16,16];
traindb='D:\matprog\matfiles\trainfiles';
filePattern=fullfile(traindb, '*.mat');
matFiles = dir(filePattern);
for i = 1:length(matFiles)
baseFileName = fullfile(traindb, matFiles(i).name);
ref_files{i} = load(baseFileName);
trainingfeatures(i,:)=extractHogfeatures(ref_files,'cellsize',CS);
end
traininglabels=traindb.Labels;
classifier=fitcecoc(trainingfeatures,traininglabels);
ERROR:
Dot indexing is not supported for variables of this type.
how to give labels and is it correct way to follow?

采纳的回答

Pratyush Roy
Pratyush Roy 2021-11-2
Hi Santhosh,
The traindb variable in the code refers to a string and not a struct. Hence the dot indexing does not work in this case.
As a workaround, you can use Image DataStore to store the image data, since it supports dot indexing.
Hope this helps!

更多回答(0 个)

类别

Help CenterFile Exchange 中查找有关 Data Distribution Plots 的更多信息

Community Treasure Hunt

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

Start Hunting!

Translated by