Whale sound signal denoising based on SVMD and improved wavelet thresholding

Abstract Cetaceans have formed a set of sonar systems that rely on acoustic waves for communication, positioning, and environmental recognition in their long-term evolution. This sonar system is characterized by strong anti-interference ability, high localization accuracy, and strong recognition ability. Effective denoising of cetacean sound is the first link in the process of applying cetacean signal analysis. For the problem of effective denoising of whale sound signals in complex underwater environments, a new denoising method based on successive variational mode decomposition (SVMD) and improved wavelet thresholding is proposed. Firstly, the noisy high-frequency intrinsic mode functions (IMFs) obtained by SVMD decomposition are sieved by the correlation coefficient method; then, these high-frequency components are subjected to improved wavelet thresholding for noise reduction; finally, the signal is reconstructed with the low-frequency IMFs. The simulation results show that the denoising method works well, and the signal-to-noise ratio is high and the root-mean-square error is low, which effectively preserves the important information of the original signal.