Wavelet denoising of Gaussian peaks: A comparative study

摘要: Abstract In this paper we apply some recent results on non-linear wavelet analysis to simulated noisy signals of chemical interest. particular, compare the soft universal thresholding algorithm described by Donoho, Fourier filters and polynomial smoothers such as Savitzky-Golay (SG). All reconstruction were evaluated basis three different criteria: mean squared error (MSE) both for whole signal an interval centred around peak, signal-to-noise ratio (SNR) improvement change in peak area. The data consists narrow Gaussian peaks with white noise. Signals low SNR investigated, since is a challenging problem each filter. Four common wavelets (Haar, Daubechies, Symmlets Coiflets) selected denoising. Our show that under chosen conditions denoising (WD) gives most cases superior performance over classical filter techniques.