Photonic Data Science

A wavenumber and intensity calibration procedure of Raman spectra by using chemometric techniques is presented. This approach allows the fine tuning of calibration parameters and routines with the final goal to eliminate setup dependent differences within experimentally recorded Raman spectra. This seems to be necessary since more and more Raman databases are needed for different analytical tasks, like identification of minerals or bacteria. Minimizing the impact of the applied experimental Raman setup on the reference (database stored) Raman spectra allows the databases to be enlarged very easily by feeding the database with Raman spectra recorded with different setups. Furthermore the chemometric analysis performance increases due to the larger number and better quality of reference spectra.