How to create a database of feature vectors for images?

I have a database of around 4 classes each consistent of around 50 images.I created a features vector on matlab to extract the features i need from each image.My question is how to create the database of vectors for all the images to then train them per class .

Answer 1

[KSSV]

N = 50 ;  % number of fimages 
C = 4 ; % number of classes 

%%loop for each image 
iwant = cell(N,C) ;
for i = 1:N
    %%Extract the features of images,
    iwant{i,1} = rand(10) ;
    iwant{i,2} = rand(5) ;
    iwant{i,3} = rand(6) ;
    iwant{i,4} = rand(11) ;
end
%%Rows corresponds to each image, and columns correspon to each class of the respective image

[Elias Unk]

Hi kssv i didn't understand the code well may you explain it showing how i extracted the features from the image and placed each features vector in an erray

[KSSV]

How I would know the way you extracted the features? In place of rand you need to fill your feature vectors.

[Elias Unk]

i mean let's say my features vector is called vect will the code change as follows? %% Extract the features of images,

   iwant{i,1} = vect(10) ;
      iwant{i,2} = vect(5) ;
      iwant{i,3} = vect(6) ;
      iwant{i,4} = vect(11) ;
  end

as well as what do these numbers in vect stand for 10,5,6,11

[KSSV]

If vec1,vec2, vec3 and vec4 are your feature vectors for image 1. Use: iwant{i,1} = vec1 ; iwant{i,2} = vec2 ; iwant{i,3} = vec3 ; iwant{i,4} = vec4; end

[Elias Unk]

nope i have 1 feature vector of 110 elements i just combined all the features extraction in 1 final vector so that each image will have only 1 vector extracted of it formed of 110 elements that i then normalized and transformed to 1 element,hope that made you better understand the issue.