A New Algorithm for Synchrosqueezing: With an Application to Multicomponent Signals Sampling and Denoising

Sylvain Meignen (1) , Thomas Oberlin (1) , Steve Mclaughlin (2)

1 MGMI - Modélisation Géométrique & Multirésolution pour l'Image
2 EPS-HWU - School of Engineering and Physical Sciences [Edinburgh]

In this paper, we address the practical implementation of the synchrosqueezing method. The emphasis is placed on the importance of the wavelet choice and we propose a novel algorithm based on a two step strategy: a mode detection step followed by a reconstruction step. Simulations will illustrate how the proposed procedure compares favorably with empirical mode decomposition and other related methods in terms of mode-mixing. We conclude the paper by studying the sensitivity of the synchrosqueezing to sampling and an application to signal denoising.

Mots clés: SynchroSqueezing; Wavelets; EMD; Time-Frequency; Sampling; Denoising

Domaines: Recherche d'information [cs.IR]

Sylvain Meignen, Thomas Oberlin, Steve Mclaughlin. A New Algorithm for Synchrosqueezing: With an Application to Multicomponent Signals Sampling and Denoising. 2012. ⟨hal-00667190⟩