A Novel Ridge Detector for Nonstationary Multicomponent Signals: Development and Application to Robust Mode Retrieval
Nils Laurent, Sylvain Meignen
Abstract
Time-frequency analysis is often used to study non stationary multicomponent signals, which can be viewed as the surperposition of modes. To understand such signals, it is essential to identify the ridges associated with the modes in the time-frequency plane. As existing ridge detectors are often not enough robust to noise, we here develop a novel approach to ridge detection based on the gathering of ridge portions in the time-frequency plane, which we coin RRP-RD. Such a technique is proved to be much more robust to noise than state-of-the-art methods based on the same framework, and we also demonstrate its benefits for mode retrieval.
Topics & Concepts
RidgeDetectorComputer scienceNoise (video)Mode (computer interface)Plane (geometry)Time–frequency analysisDetection theoryArtificial intelligenceAlgorithmSpeech recognitionComputer visionTelecommunicationsMathematicsGeologyGeometryOperating systemFilter (signal processing)Image (mathematics)PaleontologySpeech and Audio ProcessingStructural Health Monitoring TechniquesTime Series Analysis and Forecasting