The use of deconvolution and iterative optimization for spectrogram interpretation

The problem of spectrogram interpretation is considered under the assumption that the parameters of spectral peaks-their positions and magnitudes-contain the information essential for spectrometric analysis. The subsequent use of Tikhonov deconvolution and iterative correction of the estimates of those parameters is proposed. Deconvolution is used for transforming the processed spectrogram in such a way as to facilitate finding initial estimates of its parameters. The advantages of the proposed approach, i.e., gains in accuracy of estimating the parameters of peaks, are demonstrated using both synthetic and real-world spectrophotometric data