A New Ridge Detector Localizing Strong Interference in Multicomponent Signals in the Time-Frequency Plane

Sylvain Meignen (1) , Marcelo Colominas (2)

1 DAO - Données, Apprentissage et Optimisation
2 IBB Institute for Research and Development in Bioengineering and Bioinformatics

In this paper, we define a new ridge detector that enables to localize strong interference in multicomponent signals in the time-frequency (TF) plane. Each mode of a multicomponent signal can usually be associated with a ridge in the TF plane, but this is no longer the case when strong interferences occur in the signal. The new ridge detector we propose is thus designed to determine when such situations happen in the TF plane. We show that this knowledge helps to determine an appropriate window length in the definition of the spectrogram, as well as the nature of the strong interference detected. An application of the proposed approach to voice signals concludes the paper.

Mots clés: Time-frequency analysis; short-time Fourier transform; spectrogram ridges

Domaines: Traitement du signal et de l'image [eess.SP]

Sylvain Meignen, Marcelo Colominas. A New Ridge Detector Localizing Strong Interference in Multicomponent Signals in the Time-Frequency Plane. IEEE Transactions on Signal Processing, 2023, 71, pp.3413-3425. ⟨10.1109/TSP.2023.3311513⟩. ⟨hal-04218423⟩