Multicomponent signal denoising with synchrosqueezing

In this paper, we develop a new technique based on the synchrosqueezing method to denoise multicomponent signals. The approach proposed is based on a two step strategy: a mode detection step followed by a reconstruction one. The emphasis is put on the robustness of the detection step in a noisy context, a key issue in the implementation of the method. Numerical applications show the improvement in terms of multicomponent signal denoising brought about by the proposed method over the translation-invariant wavelet thresholding.