[Performance analysis of threshold denoising via different kinds of mother wavelets]

An ideal spectrum signal prototype is constructed in this paper based on the infrared ray spectrum of octane levelmeasurement to evaluate the performances of wavelet based threshold denoising approaches via different combinations of mother wavelet functions and thresholds. A performance index eta is defined to assess the signal-to-noise ratios (SNR) of denoising results, inconsideration of the trade-off between the SNR and the distortion of the original signal after wavelet denoising. Three families of mother wavelets (Symlets, Daubechies and Coiflet), four threshold selection rules (Rigrsure, Sqtwolog, Heursure and Manimaxi), and three threshold rescaling methods (One, Sln and Mln) are tested in a series of experiments to estimate the functioning of those wavelets and thresholding parameters. Experimental results show that in the cases investigated in this paper, the best denoising performance is reached via the combinations of Daubechies9 or Symlet7, 11, 14, 15 wavelets, "Rigrsure" threshold selection rule, and "Sln" threshold rescaling method.