EMD-based extraction of modulated cavitation noise

The empirical mode decomposition (EMD) is applied to extract information of modulation from signals contaminated by noise. The EMD method is capable of recovering the amplitude-modulated components from strong background noise in an adaptive way, and achieves better performance than traditional methods. We further propose a modified EMD technique by estimating the local mean of a signal via windowed average. This method alleviates the unfavorable influence of noise disturbance effectively in the process of sifting and yields a remarkable improvement for the modulation extraction. Finally, by utilizing this novel technique, we achieve the adaptive and effective extraction of the modulated cavitation noise from ship-radiated noise.