Comparison of wavelet transform and Fourier self-deconvolution (FSD) and wavelet FSD for curve fitting

Abstract

The advantages of combining Fourier self-deconvolution (FSD) and wavelet transform with curve fitting for the analysis of severely overlapped bands are compared. It is shown that, for overlapped peaks with lower signal-to-noise ratios (SNR), the method of combined wavelet transform with curve fitting provides significantly better results. In contrast, the method of combined FSD with curve fitting shows better results for severely overlapped peaks with higher SNR. Consequently, when wavelet-FSD, which is based on the combination of wavelet transform and FSD, is used to resolve severely overlapped peaks prior to curve fitting, it is shown that there is a great improvement in the conditioning of curve fitting even for severely overlapped peaks with higher noise levels. Therefore, more accurate peak parameters are achieved.