I have folder with name 'images' within that folder 20 sub-folders are there. that sub-folder consist 20 images with name like 'img (1).gif'. I have to read images one by one then store the hog features in the text file.
1 visualización (últimos 30 días)
Mostrar comentarios más antiguos
bamini thavarajah
el 29 de Ag. de 2017
Comentada: Cam Salzberger
el 31 de Ag. de 2017
objects={'apple','bat','beetle'};%apple,bat,beetle are sub-folder name
trimgs=10;
for i=1:length(objects)
for j=1:trimgs
imgpath=['.\images\' objects{i} '\img (' int2str(j) ').gif'];
I=imread(imgpath);
J = imresize(I,[256,256]);
fid=fopen('feature_vector.txt','w');
fprintf(fid,'%i\t',hog_feature_vector(J));
fprintf(fid,'\n');
fclose(fid);
end
end
0 comentarios
Respuesta aceptada
Cam Salzberger
el 29 de Ag. de 2017
What issues are you running into? Or what is your question?
I can suggest that you consider using "dir" to automatically get the "objects" name.
If you are using the same "feature_vector.txt" file for all the images, you probably want to use fopen with the 'a' option (for "append") rather than the 'w' option (for "write", discarding current contents).
2 comentarios
Cam Salzberger
el 31 de Ag. de 2017
I don't know what "hog_feature_vector". However, if it is a vector of integers, then your line:
fprintf(fid,'%i\t',hog_feature_vector(J));
seems like it would work correctly. What are you seeing when you run this? I would expect it to output something similar:
x = [1 2 3 4 5];
fprintf('%i\t',x)
1 2 3 4 5
If by "matrix output" you mean you want square brackets and stuff, you can just add that before you print each line.
Más respuestas (1)
Image Analyst
el 29 de Ag. de 2017
2 comentarios
Image Analyst
el 30 de Ag. de 2017
The link you gave just points back here. Basically learn how to use fprintf() and you can write out anything you want to a text file.
Ver también
Categorías
Más información sobre Get Started with MATLAB en Help Center y File Exchange.
Productos
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!