MATLAB Answers

Soum
0

ERROR From STD 'standard deviation'

Asked by Soum
on 10 Jul 2013
Hello every one; is anyone knows how can I extract the error between two images from their variance? knowing that I have the STD which equals = SQRT(Variance)
Thanks ;

  0 Comments

Sign in to comment.

1 Answer

Answer by Image Analyst
on 10 Jul 2013
 Accepted Answer

If you define error as the standard deviation of the absolute value of the difference between the image intensities, then you can do
differenceImage = abs(double(image1) - double(image2));
theErros = std(differenceImage(:));
You might also look at other definitions, like MSE ( http://en.wikipedia.org/wiki/Mean_squared_error), PSNR ( http://en.wikipedia.org/wiki/PSNR), and SSIM ( http://en.wikipedia.org/wiki/SSIM).

  5 Comments

Soum
on 10 Jul 2013
In my case the ERROR=(Noisy Image-De-noised Image), but I know the SDT and I wanna get the ERROR ok??
You cannot get a difference image, (Noisy Image-De-noised Image), from a single number, like 6.54868543 for example. I hope you understand why. It seems so basic that I shouldn't even have to explain why. What if I told you the mean of an image? Could you tell me what the original image looked like? No, of course not, because there are an infinite number of images that could have that mean. Well it's the same with the standard deviation - there's just not enough information to recover the original image(s).
Soum
on 11 Jul 2013
mmm ok Thanks

Sign in to comment.