Hilbert-Huang Transform(HHT) is an efficient adaptive analysis method in nonlinear and nonstationary signal analysis, processing and feature extraction, which is widely used in engineering field. In this paper, Empirical Mode Decomposition(EMD) and Variational Mode Decomposition(VMD) are employed to compare and analyze the underwater acoustic signals of offshore motorboats. It is found that the energy of underwater acoustic signal is mainly concentrated in the low frequency band, and its amplitude is relatively large compared with that in the high frequency band. When analyzing this kind of signal, mode mixing is produced by EMD method, therefore using EMD cannot effectively decompose the signal and extract features. Nevertheless, using VMD method can effectively reduce the phenomenon of mode mixing and successfully extract its signal characteristics. The results show that VMD method is more effective in ship underwater acoustic signal processing and feature extraction.