RMS image process for multiple photographs

I need the RMS result of about 100 pictures like above. First of all, I want to bring the image, use the 'rgb2gray' code, use the 'RMS' code, and add the thresholding value to dichotomize the picture. After all, it is my goal to observe the boundaries between the background and the subject clearly. Please help me with this problem.

Answer 1

[Image Analyst]

What is the RMS result and RMS code? What does dichotomize mean? You mean binarize or threshold???

What I'd do is, if you want the average of all the spray cone patterns (see the FAQ) is to cast the images to double and sum them up. Then I'd threshold the sum image and call bwboundaries. You might find my interactive thresholding app useful: https://www.mathworks.com/matlabcentral/fileexchange/29372-thresholding-an-image

[JCH]

Thank you for your reply. Let me give you some additional information. Please check my information and give me an additional reply.

First, The RMS Eq is as follow.

The RMS method of taking the root-mean-square (RMS) average image was used throughout the injection period to compare the spray shapes. This method was used to determine the spray shape and to produce high-contrast images. However, it is difficult to produce high contrast images using the mean averaging method due to the spray shape. The RMS allows for easier investigation of the injection area and spray dispersion angle. The RMS average was used to produce full-resolution averages of an ensemble of frames using that Eq.

And, this Figure shows examples of various averaged images in which the RMS image produces a clear boundary.

Ref : https://www.sciencedirect.com/science/article/pii/S0016236111007721

[Image Analyst]

OK, easy enough if you have all the S{k}, but how do you get those? I'm sure you have all the raw images but how do they say they "correct" and "denoise" the images before squaring them?

If you just want to find the shape, and not analyze any pattern inside the shape, another option to look at is just to take the max:

for k = 1 : numImages
    outputImageMax = max(cat(3, outputImage, S{k}));
    outputImageRMS = outputImageRMS + double(S{k}) .^ 2;
end
% Need to divide the RMS image (only) by the number of images.
outputImageRMS = outputImageRMS / numImages;

Since it's so easy, do it both ways and look at the images.

[JCH]

Okay. It was a great help for me. Thank you so much.