hi i have mean and standard deviation in one image now i want to find entropy for RGB color and skewness..

how to find entropy and skewness for RGB in one image?

Answer 1

[Youssef Khmou]

To find the entropy, you have two ways, using predefined function or by programming :

 H=imread('autumn.tif'); % RGB sample
 E1=entropy(H)

Using program here is way :

 H=im2double(H(:));
 P=hist(H,length(H));
 P(P==0)=[];
 P=P/sum(P);
 E2=-sum(P.*log2(P));

[zack]

thanks..RGB sample mean we need to change this picture to each color(RGB)and test this code for each RGB or just take the picture that have color....and how about skewness..any idea for this code?

[Youssef Khmou]

You can handle the entropy using different ways, you can average for each channel :

 H=imread('autumn.tif');
 ER=entropy(H(:,:,1)); % Red
 EG=entropy(H(:,:,2)); % Green
 EB=entropy(H(:,:,3)); % Blue

The average is

 E=(ER+EG+EB)/3

The same concepts can be used for skewness :

 SK=skewness(im2double(H(:)));

[zack]

i see..so its same with the code that u give before?..let say im use predefined fucntion, i need extract 3 color channel RGB to test for each color?..what is different with new one..its run simulataneously..

[zack]

do u know skewness using program?

[Youssef Khmou]

It is possible to use program, map the matrix into a vector r and compute the skewness :

mean( ((r-mean(r))/std(r)).^3)

[Image Analyst]

zach, not sure what that means, but if you mean to compute it yourself, did you look at the program I posted for you where one of the lines is:

% Get the skew.
skew = sum((GLs - meanGL) .^ 3 .* pixelCounts) / ((numberOfPixels - 1) * stdDev^3);

[zack]

can u gv me details about variable r..its the row from which one matrix of pixel.and what should i put for pixelcounts and numberofpixel from my pixel image.

[Youssef Khmou]

 H=imread('autumn.tif');
 r=im2double(H(:));

[zack]

awesome..tqvm..:)

Answer 2

[Image Analyst]

For entropy you can use entropy(). If you want a local entropy to give an image of how much the entropy is around the image, you can use entropyfilt(). If you want image moments, see the attached demo of mine.

A color image is three color channels. You can take the entropy of each.

% Extract the individual red, green, and blue color channels.
redChannel = rgbImage(:, :, 1);
greenChannel = rgbImage(:, :, 2);
blueChannel = rgbImage(:, :, 3);

redEntropy = entropy(redChannel);
greenEntropy = entropy(greenChannel);
blueEntropy = entropy(blueChannel);

[zack]

image moments mean its only take entropy for RGB only?because i dont need in grayscale..i have see ur attached file but cannot find for entropy code..

[zack]

thanks..may i know from this code skew = sum((GLs - meanGL) .^ 3 .* pixelCounts) / ((numberOfPixels - 1) * stdDev^3);,where can i get pixelcounts and number of pixel from my image pixel..and i can still have no idea to get entropy for RGB..appreciate ur help.:)

[Image Analyst]

Was this comment posted before you used Youssef's code and accepted it? He only did entropy and you accepted it so I assume that it does what you need.