Can someone tell me how to find the summation of all the pixel values in an image?

I'm trying to implement a Sum Average Difference (SAD) based code to find the difference between consecutive frames in a video to eliminate those frames which do not contain activity i.e. no motion.

[Bruno Pop-Stefanov]

Do you mean the sum of absolute differences?

[Sneheet]

yes, the absolute values of pixel differences.

Answer 1

[Bruno Pop-Stefanov]

Assuming the frames you want to compare are called img_prev and img_next, both matrices of the same size, the following function computes what you want:

function out = sad(img_prev, img_next)

% First, take the absolute value of the difference at each pixel
myAbsDiff = abs(img_prev - img_next);

% Then, sum over all pixels
out = sum(myAbsDiff(:));

end

Using (:) you can transform matrix myAbsDiff into a vector. Calling sum on a vector computes the sum of all elements in a vector.

[Sneheet]

sad is an in-built function?

[Image Analyst]

No, as you can see he just defined it right there above. However, with uint8 images, it will need to be modified to allow negative numbers by casting the numbers to double.

function out = sad(img_prev, img_next)

% First, take the absolute value of the difference at each pixel
myAbsDiff = abs(double(img_prev) - double(img_next));

% Then, sum over all pixels
out = sum(myAbsDiff(:));

end

[Walter Roberson]

No, Bruno has given you the code for it.