Sparse data in surface plots

[Thomas Klingler]

I am using surface plots and sometimes have missing grid data points.
I'd like to interpolate or curve fit to fill in the missing points in the surface.
Any ideas on methods for doing this ?

Thanks,
Later,
Paul

[Narcís]

Hi Paul,

You can find an example for that at the TeeChart features demo, available at its program group. The example can be found at What's New?\New in Series\Surface Speed\Surface Smoothing.

[Thomas Klingler]

Thanks for the reply.

I don't really need to add more points, but want to fill in null values in the surface based on surrounding surface points. Now (in V7.01) undefined surface points show up as clear 'holes'.

Any support for that ?

Thanks,
Later,
Paul

[Pep]

Hi Paul,

one thing you can do is to iterate through all the data added checking which are the missing points and then add them as extra data to the Series.

[Thomas Klingler]

Yes, I'd love to add in the missing data, and that is what my basic question is about ... how to calculate / interpolate the missing data based on the known surface points around the missing data such that the surface appears smooth.

The surface is in fact the reconstruction of the retina of the human eye from stereo fundus images. Parts of the image surface can't be calulated due to realities of image quality and artifacts.

I know there are techniques called 'cubic b-spline' interpolation that can calculate missing 3d data from known control points.

So my questions is, does anyone have an implimentation of that or any other technique that works within the context of Tchart surface plots ?

Thanks in advance for any info.

Later,
Paul

[leaf]

There are many of these routines on the internet. I would suggest starting on the Numerical Recipes website. They have a few different interpolation routines.