How to plot gaussian curve for each data point of boundary ?

[Mathieu NOE]

I don't understand the question - what is your image about ?

[Charu]

Its an extracted boundary from an image . Actually i want to calculate RMSF (ROOT MEAN SQUARE FLUCTUATION ) at each point between two images (deviated one and non deviated one)..and want to plot as pdf (probability density function) as gaussian curves with the calculation of skewness and kurtosis.

[Mann Baidi]

It would be helpful if you can describe more about the issue

Answer 1

[Image Analyst]

If you have the Statistics and Machine Learning Toolbox, what I'd try is pdist2 and put in the boundaries of both curves. Then for each row you can find out which two points are closest to each other. Then get the average of those closest distances. Something like (untested)

distanceMatrix = pdist2(xyRef, xyTest);
[rows, columns] = size(distanceMatrix)
% Allocate arrays for closest distances.
closestColDistances = nan(1, columns);
closestRowDistances = nan(1, rows);
% Get rid of zeros (distance of point to itself)
distanceMatrix(distanceMatrix==0) = inf; % Set to infinity.
% Get distances of ref curve to test curve.
for row = 1 : rows
    closestRowDistances = min(distanceMatrix(row, :));
end
% Get distances of test curve to ref curve.
for col = 1 : columns
    closestColDistances = min(distanceMatrix(:, col));
end
% Get the average distance.
averageDistance = mean([closestRowDistances, closestColDistances])

You need to get the distances both ways because if you only look at, say, the distance from the ref to the test, there may be some points on the test curve that are not closest and then they would not contribute to the average distance.