How to apply a Gaussian filter to co-ordinate data

I am trying to measure the surface roughness of a component. I have the raw trace of the surface but in order to extract just the roughness I need to filter out the longer wavelengths of form and waviness using a Gaussian filter.

I have 32,000 co-ordinate points to filter through. The transition point between waviness and roughness needs to be left as an independent variable.

Can anyone help me with the code for this?...

Answer 1

[Image Analyst]

You can use conv2() or imfilter() with a large kernel, then subtract that from your original image. Or you can do it in the fourier domain. I have attached a Fourier example.