VI to compute partial derivatives and curvature on a surface

[WCR]

I have a surface ( a set of x,y,z triples that form a grid, from measurements, not from a formula).  I want to compute the curvature at each point on the grid.  I have already done some smoothing  and interpolation, so it does not have any really sharp edges.  I understand that I can't compute curvature all the way to the edge since I need adjacent points.  Is there a built-in LV routine for this, or must I write my own? 

If I must write my own routine, I will write a nested loop that goes along each row and column of the grid.  At each point, I will estimate the matrix of 2nd partial derivatives by finite differences.  From that I will compute the maximum curavature, or minimum radius of curvature, at that point (once I figure out the formula, which I have not quite done).

I appreciate any suggestions; I hope there's already a VI to do this. Thank you.

[Aldo_A · Accepted solution]

Hi WCR,

For curvature computation, we only have an IMAQ VI called "IMAQ Classify Curvature.vi" (http://zone.ni.com/reference/en-XX/help/370281P-01/imaqvision/imaq_classify_curvature/) that will compute the curvature for you. However, it is an NI-IMAQ driver function and it only accepts a contour image. So if you are not planning to process this with a picture, then you are going to have to compute min/max curvature with our artihmetic vi's.

There is a VI which will compute the partial derivatives called "Partial Derivatives of f(x1,x2) VI" (http://zone.ni.com/reference/en-XX/help/371361H-01/gmath/partial_derivatives_fx1x2/). This particular VI will calculate a 2D array of two independent variables.

Hope this helps!

[WCR]

Thank you Aldo.  Our data is not in the form of an image and not really convertible to an image.  Thanks for revealing the existence of a Calculus VIs palette.  It computes partials based on a formula.  My data is experimental, so I don't have a formula for it, and I think the Partials VI will not help me.

I will write my own VI, using some suggestions in "Fast Algorithms for Ridge Detection" by Liang et al, 

Proceedings of 1997 International Conference on Shape Modeling and Applications, IEEE Comput. Soc. Press.

Bill