How to calculate the average gradient of an image?

i have tried this code.But the results are mismatching.

[Gmag, Gdir] = imgradient(X, 'sobel');

AG=abs((sum(sum(mode(Gmag))/(sqrt(2)))))

[Raunak Gupta]

Hi,

Can you provide from what you are comparing with, while generating the result?

[Munshida P]

original image and enhanced image(DICOM format)

[Munshida P]

Sir, i can't attach my dicom image here because,the file format is not supported.(.dcm)

Answer 1

[Guillaume]

Missing from your equation, is the exact definition of G. You're currently using the sobel operator to compute the gradient and as documented in the algorithms section of imgradient whichever operator you use is going to give you vastly different results for the magnitude.

I suspect that your equation use a completely different definition than 'sobel' for G.

As asked by Raunak, for us to be able to tell you what is wrong, we need the actual source image and whatever you compare it to.

Answer 2

[Bruno Luong]

AG=abs((sum(sum(mode(Gmag))/(sqrt(2)))))

I can't see the denominator factor (H-1)*(W-1). And Why you use MODE?

Those two errors explain discrepency already.

[Munshida P]

|Gmag| is there in the equation.....so i used mode function

[Bruno Luong]

There is no relation between |Gmag| and MODE.

If you code arbitrary without to trying understand what the formula means and what the function does, then you'll get arbitrary result.

[Munshida P]

|Gmag| means?

[Guillaume]

is the absolute value of x, nothing to do with the mode (particularly as mode(x) = x for scalar x). As Bruno says, you need to understand your formula before trying to implement it.

In the same respect, you need to know how G(x, y) is calculated. As I answered, it's probably not calculated with the sobel operator which will give different results to other gradient operators. Wherever you found that formula, there should also be a definition for G(x, y).

edit: in fact, you can be guaranteed that G is not imgradient since imgradient will always return a positive gradient magnitude, so there would be no point in the taking the absolute value.

[Munshida P]

k.Iam new to matlab.need more to study about it.i will try once more

[Guillaume]

So far, the issue does not appear to be matlab but maths.

[Munshida P]

k

Answer 3

[Raunak Gupta]

Hi,

As per the formula you have mentioned above, the syntax for Calculating ‘AG’ is mistaken. For implementing the formula, you can use below commands.

%Here X is the array for which Gradient is to be calculated
[Gmag,Gdir] = imgradient(X,'sobel');
% Here since Gmag represent Magnitude of Gradient so no need to take absolute value
AG = sum(sum(Gmag))./(sqrt(2)*(size(X,1)-1)*(size(X,2)-1));

If you want to try other method for calculating gradient you may investigate ’method’ parameter in imgradient. Please keep in mind that imgradient works only for grayscale images and would not provide required results for RGB images. Here is the link to the documentation of imgradient.

https://www.mathworks.com/help/images/ref/imgradient.html

[Munshida P]

k...i will try this code

[PPO POT]

Dear Munshida P,

Please tell me about this research paper title,

I will be investigate this problem.

[Munshida P]

base paper:"A new approach for medical image enhancement through luminance level modulation and gradient modulation"

[PPO POT]

Thank you, Munshida P

[PPO POT]

https://www.sciencedirect.com/science/article/pii/S1746809418302714

This is my code, with this code you can browse image file and then calculate AG automatically,

[filename1,pathname] = uigetfile('*.*',' Select the Original Image ');

orgIM = imread(filename1);

% Pattern

%[Gmag, Gdir] = imgradient(I,'prewitt');

% 1] If test image is true color covert its to grayscale

if size(orgIM,3) == 3

testIM = rgb2gray(imread(filename1));

[Gmag,Gdir] = imgradient(testIM,'prewitt');

% Calcualte Average gradient

AG = sum(sum(Gmag))./(sqrt(2)*(size(testIM,1)-1)*(size(testIM,2)-1))

end

[Munshida P]

info = dicominfo('MR-MONO2-12-an2');
X= dicomread(info);
imshow(X,[])
[Gmag,Gdir] = imgradient(X,'prewitt');
%Average gradient
AG = sum(sum(Gmag))./(sqrt(2)*(size(X,1)-1)*(size(X,2)-1 ))

got it....thank you sir