Image Processing: Change RMS keeping Mean Constant

I'm trying to find a program that will help with image processing for vision experiments. Specifically, I need the mean of the image to remain constant, but I want to change the RMS contrast to different levels. Is there a program out there to help with this?

I've tried a number of different approaches but they're not working reliably with larger contrasts.

Thanks in advance for your help!

Answer 1

[Image Analyst]

I would think that the rand() function should help you immensely.

[Emily]

Sorry, I wasn't clear. I want to read in a particular image (e.g., a face) and then scale it so it preserves the same image, same mean luminance, but changes the RMS contrast to different values.

[Image Analyst]

What is your reference image?

squaredDifferenceImage = (testImage - referenceImage) .^ 2;
meanSquaredError = mean(squaredDifferenceImage(:));
rms = sqrt(meanSquaredError);

[Emily]

A reference image? I'm just taking an image, say a face with a mean of 74 and RMS of 50, and want to change the contrast of the same face so it has a mean of 74 and an RMS of 20, for example.

Does that make more sense? Thanks for your thoughts on this.

[Image Analyst]

An RMS is composed of two signals. What are your two signals? Is one just a uniform image with the mean of the image?

referenceImage = mean(testImage) * ones(size(referenceImage));

???