How can I catch a wave and measure asymmetric peak?

A profilometer using the topology of a deposited material surface can be measured. I want to design a program to read in the raw data of these profiles files and then perform an analysis that will extract the distribution in the heights of peaks/troughs and the distances between the peaks. Reading data, manipulation of tables and producing output/plots files with that I'm fine but the problem I have are as to what analysis tools can be used on these asymmetric waveforms. I was thinking about a form any signal processing technique, but I'm not sure. I have attached a profile example for one of my samples to give an idea of what I am working with...

If it is representative enough of your data, I would first take stabbed using the detector tip a LabVIEW VI (Signal Processing-> operation of signals). There not much noise at all relative to the size of your pics (if we talk only about those visible here), so you could just smooth out your data with a Savitsky-Golay function (or such) to exit the small move of random noise, offset all of your points to make positive, set a width value and the crest of the reasonable threshold and get there.

Now, if you have a huge hollow as in the file you sent, you may miss a lot of pics (even once, I don't know what are your criteria for a 'Summit'). If that happens, I would say looking at data upside down to find your valleys as reverse 'summits '. At this point, you can go right back, connect the valleys with lines and subtract these points from your data to a pseudo-base and run the VI Ridge detector again (you probably can go much more sensitive with your adjustment of threshold because you took most of drifting in the very long term). You should have now identified all of your points of peak and Valley on axis X. return the original data to determine the heights of peaks with some basic algorithm you want to use (linear between valleys, Bézier curves, etc.).

That's how I would attack it generically. But it's a very common situation in probably all science and engineering fields. Isn't there some geological no - power of data standard method generally agreed criteria for your problem?

Cameron

Tags: NI Software

Similar Questions

Maybe you are looking for