How do I perform multiplicative scatter correction on a set of NIR data?

I have 60 near infrared (NIR) spectra that I would like to pre-treat using multiplicative scatter correction (MSC). I think I may need a specific toolbox (Statistics I think). I believe I have that already but not sure where to go from there!

[Nellikin]

I need to do this too? Sadly, no answers...

Answer 1

[Md Syduzzaman]

I am new here and not an expert, but try this, it might help. someone posted it.

Answer 2

[amengmeng]

Multiplicative Scatter Correction Algorithm As an estimate of the ideal sample, we can use for instance the average of the calibration set. Multiplicative scatter correction performs best if an offset correction is carried out first. For each sample:

where

xi: the ith spectrum of the collection used for calculation.

a, b: For each sample, a and b are estimated by ordinary least-squares regression of spectrum xi versus yAverage over the available wavelengths j.

Each value xij of the corrected spectrum xij (MSC) is calculated as:

where

xij: the intensity of the ith spectrum and jth wavelength of the collection used for calculation.

Answer 3

[M.L]

it's a POLYNOMIAL CALCULATION;

you have to correct your 60 spectra(data_spectra) compared to de reference spectra

%% calculation

coeff=zeros(size(data_spectra,2),2);

for i=1:size(data_spectra,1) % MSC (y=ax+b)

coeff(i,:) = polyfit(data_ref,data_spectra (i,:),1);

end

N.B data_ref could be one of your 60 spectra

Then you correct you spectra based on the calculated coeff

data_corrected(x) will be: for each spectrom x=data_sepctra(i,:)-coeff(i,2)/coeff(i,1) % x=y-b/a

[Merkebu Getachew]

are these scripts on R or MATLAB?