10-01-2015 11:38 AM
Hi,
I am using Laser Induced Fluorescence to identify the material. I have attached a sample of LIF in Signal.png. I am looking a algorithm to identify the material using the spectra.
amplitude of the spectra can change but the shape will almost remain same with change in the condition.
Can anyone suggest the algorithm?
10-01-2015 12:43 PM
What kind of material? If it's any kind of complex compound I would think you need a UV spectra library to search against. Not an algorithm.
10-01-2015 01:26 PM
Normalize.vi that ought to help some
10-02-2015 01:36 AM - edited 10-02-2015 01:47 AM
Some things that left from my >20a ago Raman spectrometer use: Assuming that the spectrum is built out of n Gauss peaks, you need an (maybe already existing) LabVIEW algorithm to (interativly) match n gauss peaks to your spectrum. n can be (manual) guess , or you define a error boundary and let the algorithm increase n until that max error condition is met.
Each peak is defined by wavelength location, sigma and a normalized amplitude relation. For your posted spektra n migth be 3 (or 4) so n times 3 parameters characterize your materiaL:
There are tons of papers and lectures on spectrometer data interpetation.....
10-02-2015 08:59 AM
@Henrik_Volkers wrote:
There are tons of papers and lectures on spectrometer data interpetation.....
Agreed. I used to work with IR/NIR spectrophotometers so I was intrigued by the OP's question and found a LOT of search results for analyzing spectra. We can help you with the LabVIEW part but you might want to look into some physics forums as well for help on the math and analysis parts.
Do you have a particular instrument you're using? Perhaps there are LabVIEW drivers or some pre-existing software for it that you could leverage.