Getting divergence of a 3D pointcloud

Hi All,

I am trying to get the divergence values of a 3D pointcloud scan and I am having troube fitting the data to the "divergence" function. I have a sample of position and vector values:

pos=[
76.8777310000000    48.4557610000000    -64.5356670000000
57.6735570000000    57.4872930000000    -68.8201290000000
6.32152900000000    65.5864330000000    -66.0213850000000
79.1429290000000    50.6963460000000    -69.9525910000000
63.6980360000000    56.9969980000000    -70.2472530000000
59.7197610000000    60.1693840000000    -74.0278550000000]

vec=[
0.394940000000000   0.737656000000000   0.544422000000000
0.202651000000000   0.833657000000000   0.500410000000000
0.0167880000000000  0.808301000000000   0.561047000000000
0.424063000000000   0.802266000000000   0.414878000000000
0.247707000000000   0.809888000000000   0.526833000000000
0.151668000000000   0.861146000000000   0.480130000000000]

x = pos(:,1);   %extract the first column of data
y = pos(:,2);   %extract the second column of data
z = pos(:,3);   %extract the third column of data
u = vec(:,1);
v = vec(:,2);
w = vec(:,3);

Then when I try:

div = divergence(x,y,z,u,v,w);

I get the following error:

??? Error using ==> divergence at 55 U,V,W must all be a 3D arrays.

Could someone please help, it's probably my silly formatting problem, but I cannot seem to figure it out. Much Appreciated.

[Andrei Bobrov]

please read about function 'divergence' in MATLAB help: http://www.mathworks.com/help/techdoc/ref/divergence.html

Answer 1

[Andrew Newell]

You have worse than a formatting problem, if your point cloud is, as the name implies, an irregularly spaced collection of points. The function divergence assumes that your points lie on a 3D grid. You will need to interpolate both pos and vec onto a grid before you can use divergence. Then you will need to use meshgrid to turn each coordinate x, u, etc. into a 3D array.