How to Calculate mse?

How to Calculate mse? This is correct mse = mean(mean((originalImage-embeddedImage).^2));

Answer 1

[Image Analyst]

No, only if they're floating point images, not integer images like uint8. You need to cast them to double so that your numbers don't get clipped at 0 if they would go negative.

You're best off using the built-in immse(). Otherwise if you want to do it yourself because you're missing the toolbox, see my attached code.

[Image Analyst]

Reem's "Answer" moved here:

function [PSNR]=psnr(originalImage,embeddedImage)
mse = mean(mean((originalImage-embeddedImage).^2));
MAXI=255;  %MAXI is the maximum possible pixel value of the image. 
%When the pixels are represented using 8 bits per sample, this is 255.
PSNR=10*log10(MAXI^2/mse);
end

This code

[Image Analyst]

Not if they're typical images. Like I explained and showed you in the psnr.m file I attached for you, you need to case to double if they're uint8 which most images are.

mse = mean2((double(originalImage) - double(embeddedImage)).^2);