How to find local peaks in a 3D plane

Hello all,

I have a problem in automatically detection of 4 quite obviously present peaks. I transformed my data to a almost flat plane with four peaks in z-direction. I want to automatically detect the locations and corresponding datapoints of the 4 peaks. Does someone know how to solve this problem? I've added my data.

[dpb]

You could help by supplying the code to do what you've done so far and by telling us what, specifically is the data in the .mat file -- original, transformed by whatever it is you described, ...???

Answer 1

[Akira Agata]

How about the following?

% Load data
load('positionsplane.mat');

% Arrange it as a table variable
tData = array2table(positionsplanenew,'VariableNames',{'x','y','z'});

% Extract z > 15
idx = tData.z > 15;
tData2 = tData(idx,:);

% Apply k-means clustering
rng('default');
g = kmeans(tData2{:,:},4);
tData2.group = g;

% Find maximam points for each cluster
peaks = splitapply(@max, tData2.z, tData2.group);
idx = ismember(tData2.z, peaks);

% Visualize the result
figure
scatter3(tData.x, tData.y, tData.z, '.')
hold on
scatter3(tData2.x(idx), tData2.y(idx), tData2.z(idx),'r','filled')