Hi Saurav,
There is no exisiting function in MATLAB as of now (R2024a) that can find the gradient change of a point in 3D space with respect to its neighbour.
However, you can manually do it by calculating Normals for each vertex (points) and iteratively calculate the gradient change.
MATLAB's "patch" function can be used to compute normals.
Here is how you can approch the problem:
Note: I was not able to open your STL file
% Using the stl file from the below example
% openExample('matlab/ReadTriangulationFromSTLTextFileExample')
TR = stlread('tristltext.stl');
F = TR.ConnectivityList; % Faces or Connectivity List
V = TR.Points; % Vertex or Points
p = patch('Faces', F, 'Vertices', V);
% storing the normals in a matrix
normals = p.VertexNormals;
% Initialize matrix to hold gradient changes
gradientChange = zeros(size(V, 1), 1);
% Loop through each vertex
for i = 1:size(V, 1)
% Find faces that include this vertex (row indices of a vertix in F)
[row, ~] = find(F == i);
% Find unique vertices connected to those faces, excluding the current vertex
neighbors = unique(F(row,:));
neighbors(neighbors == i) = [];
% Calculate the difference between the current vertex and its neighbors
% Here, we simply calculate the Euclidean distance as a proxy for gradient change
diffs = sqrt(sum((V(i,:) - V(neighbors,:)).^2, 2));
% Store the mean difference (or choose another metric)
gradientChange(i) = mean(diffs);
end
disp(gradientChange);
Gradient change for each point/ vertex is stored in the gradientChange matrix.
