homogenization of a point cloud

[nollMarvin]

Hello together,

i am looking for an alogrithm for computing the homogeneity of a 2D-point cloud. I want to know if the points are evenly distributed over a certain area or if there are clusters/accumulations of points.

Do you have an idea how to approach this?

greetings

M. Noll

[Henrik_Volkers]

both dimensions must be homogenious -> chi square test

and no correlation ???

make a 2D histogram (bin the values in 'meaningful' sizes), square sum the diff to mean?

[nollMarvin]

I'm no sure if i get you right, but by chi qauare test and 'square sum the diff to mean'(i.e standard deviation) you mean two different things or the same.

My first aproach was quite simmilar to what you said. I layed a grid over my test area and counted the points for each sector. then I ended up with an 2d-array of #/sector. The Standard deviation should give me a value how good the homogeneity is?

Thanks alot!

[Christian-P]

Hi Marvin,

if you're interested in the chi square test, have a look at the following VI:

http://zone.ni.com/reference/en-XX/help/371361L-01/gmath/contingency_table/

Best regards

Christian