how to find mean absolute deviation image?

I want to display mean absolute deviation image from a given input jpg Image. MAD is an energy measure that represents regularity of texture.

Answer 1

[Guillaume]

I don't see what's difficult about that. Assuming an greyscale image:

mad = mean(abs(yourimage(:) - mean(yourimage(:))))

Answer 2

[Image Analyst]

Cast to double, subtract, then get the mean:

mad = mean2(abs(double(image1) - double(image2)));

mad is a scalar. If you want the IMAGE then don't use mean2():

mad = abs(double(image1) - double(image2));

You need to cast to double so that you don't clip to zero if they're unsigned integer images.

[Khushal Gumgaonkar]

Hi, I have single Image and I want to display MAD of that. In that case what is your image1 and image2.

Thanks and Regards Khushal

[Guillaume]

Perhaps you need to give a more formal definition of what the mean absolute deviation is for you.

I've given you an answer that follows this definition for the mean absolute deviation around the mean. This assumes a single image and gives you a scalar.

Image Analyst uses a different definition that requires two different images and gives you either a scalar or an image.

[Khushal Gumgaonkar]

actually I am doing brain tumour segmentation and MAD is one of the feature. I have applied SWT(Stationary Wavelet Transform) and ordered filtering. Now I want to calculate MAD and display MAD image. Now please tell me which is image1 and image2. Thanks and Regards Khushal

[Guillaume]

I'll repeat:

give a formal definition of what the mean absolute deviation is for you

Image Analyst's answer computes the mean of the absolute deviation between two images

My answer computes the mean absolute deviation around the mean of a single image

What you want, an image from a single image, does not appear to match either definition.

[Image Analyst]

I agree with Guillaume. Explain EXACTLY what an "MAD image" is.