Quite a craftmanship is required to obtain what I call "clear" signal.
What I do:
- remove cosmic rays from images
- extract cell & background signal from image's pixels
- correct those by MSC (multiplicative Scattering Correction)
- subtract background from cell signal
- interpolate the axis (which means basically calculate a new axis) using 3 peaks of reference - because when doing experiments at different days you may have a shift in Raman x axis (Raman shift) or wave number due to optics and laser.
- perform baseline subtraction by Polyfit or other method (check google & literature)
- normalization by vector norm.
Good luck ;)
Actually, because this processing issue is recurrent, I feel I should publish my method somewhere^^
Use exponential baseline correction for raman data. If you are using origin software then it is quite easy just insert exponential while doing baseline correction. Rest details can be checked out in this video tutorial - https://www.youtube.com/watch?v=3ij7blBQ0SQ