A Multi-target Underwater Acoustic Signals Denoising Method Based on Wavelet

The actual underwater acoustic signals consist of multiple components such as ship-radiated noise and ambient sea noise. However, existing noise reduction methods are not effective for multi-target underwater acoustic signals. Aiming at this issue, an improved method which is based on wavelet threshold denoising is proposed. First of all, to obtain the wavelet coefficients, the original signal is decomposed using wavelet. After that, on account of the shortcomings of the traditional wavelet threshold denoising, an improvement way to calculate the threshold and a new wavelet threshold function are adopted to process the coefficients which have been gotten. Finally, the signal after denoising is obtained by reconstructing the new wavelet coefficients. Experimental results indicate that, the SNR of signals can be increased to above 67dB, and simultaneously the RMSE can be lower than 0.01. From the perspective of these two methods of assessment, the proposed method is superior to other methods on the noise reduction. As a consequence, in the field of underwater acoustic signals denoising, it will have bright prospects.