Method for correcting spectral data for data due to the spectral measurement process itself and estimating unknown property and/or composition data of a sample using such method
First Claim
1. A method of estimating unknown property and/or composition data of a sample, comprising:
- (i) collecting respective spectra of n calibration samples, the spectra being quantified at f discrete frequencies (or wavelengths) and forming a matrix X of dimension f by n;
(ii) producing a correction matrix Um of dimension f by m comprising m digitised correction spectra at said discrete frequencies f, said correction spectra simulating data arising from the measurement process itself;
(iii) orthogonalising X with respect to Um to produce a corrected spectra matrix Xc whose spectra are each orthogonal to all the spectra in Um ;
(iv) collecting c property data, composition data, or property and, composition data for each of the n calibration samples to form a matrix Y of dimension n by c (c>
1);
(v) determining a predictive model correlating the elements of matrix Y to those of matrix Xc ;
(vi) measuring the spectrum of the sample under consideration at said f discrete frequencies to form a matrix of dimension f by 1; and
(vii) estimating the unknown property and/or composition data of the sample under consideration from its measured spectrum using the predictive model.
1 Assignment
0 Petitions
Accused Products
Abstract
For correcting measured spectral data of n samples for data due to the measurement process itself, e.g. due to spectral baseline variations and/or water vapor and carbon dioxide present in the atmosphere of the spectrometer used to make the spectral measurements, the spectral being quantified at f discrete frequencies to produce a matrix X (of dimension f by n) of calibration data, matrix X is orthogonalized with respect to a correction matrix Um of dimension f by m comprising m quantified correction spectra, at the discrete frequencies f, which simulate data arising from the measurement process itself. The correction method is preferably included in a method of estimating unknown property and/or composition data of a sample under consideration, in which the n samples are calibration samples and a predictive model is developed interrelating known property and composition data of the calibration samples to their spectral data corrected for the data due to the measurement process itself. Then, the unknown property and/or composition data of the example under consideration is estimated from the predictive model on the basis of its measured spectrum.
287 Citations
20 Claims
-
1. A method of estimating unknown property and/or composition data of a sample, comprising:
-
(i) collecting respective spectra of n calibration samples, the spectra being quantified at f discrete frequencies (or wavelengths) and forming a matrix X of dimension f by n; (ii) producing a correction matrix Um of dimension f by m comprising m digitised correction spectra at said discrete frequencies f, said correction spectra simulating data arising from the measurement process itself; (iii) orthogonalising X with respect to Um to produce a corrected spectra matrix Xc whose spectra are each orthogonal to all the spectra in Um ; (iv) collecting c property data, composition data, or property and, composition data for each of the n calibration samples to form a matrix Y of dimension n by c (c>
1);(v) determining a predictive model correlating the elements of matrix Y to those of matrix Xc ; (vi) measuring the spectrum of the sample under consideration at said f discrete frequencies to form a matrix of dimension f by 1; and (vii) estimating the unknown property and/or composition data of the sample under consideration from its measured spectrum using the predictive model. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
Specification