How to create a features vector after extracting them from an Image?

I extracted color features like mean,standard deviation,variance etc...note that i extracted from different color spaces like RGB,HSV,YCbCr and extracted the same features for each color plane,R,G,B,H,...etc.. and texture features like correlation,contrast,energy etc..note that i used GLCM to extract my texture features and i did that for 4 orientations so some of my texture features is a 1D 4 elements vector. How can i create 1 final vector that summarizes everything for an image and which way/ways could be used so i can actually find out what features matter most so i can increase their weights accordingly or get rid of the least important.

Answer 1

[Image Analyst]

String them all together:

featureVector = [var1, var2, var3, ..... etc.]

It might help to normalize all variables to the range 0-1 so that one is not overly influential just because it has a higher value.

[Elias Unk]

Should i normalize by dividing the each element to it's norm which is the square root of the sum of the square of each element or do you recommmend any in-built functions. then should i use "cat" to group them. And finally any idea how i know which features contribute most in a successful classification.

[Image Analyst]

You might use mat2gray() which normalized arrays to the range 0-1. You can use cat() if you want. I'm a little fuzzy on how to determine "importance" of the various features - that's a stats question. I remember doing step-wise regression once is the Mathworks stats toolbox class to determine which features are really needed, but I don't have the coursebook with me and I didn't remember. I'm sure there's some kind of function in the stats toolbox to figure it out though. Take a look at the stepwise() function in the Statistics and Machine Learning help.

[Elias Unk]

so i took this example

tb=[22.9 30.0 30.3 27.8 24.1 28.2 26.4 12.6 39.7 38.0];
normalized_V = tb/norm(tb);
I = mat2gray(tb);

to check the difference results i'll get and for normalized_v i got 0.2503 0.3280 0.3312 0.3039 0.2635 0.3083 0.2886 0.1377 0.4340 0.4154

for I i got 0.3801 0.6421 0.6531 0.5609 0.4244 0.5756 0.5092 0 1.0000 0.9373 which one should i use and why

[Image Analyst]

They do different things. Either might be okay. The thing is you want to avoid values in your feature vector that are like orders of magnitude different from each other.