Feature extraction and classification for underwater target signals based on Hilbert-Huang transform theory

Abstract

In order to realize feature extraction and classification for underwater target signals, instead of the empirical mode decomposition (EMD) method, a new Ensemble EMD (EEMD) is used in the Hilbert-Huang transform when underwater target signals are analyzed. By the EEMD and Hilbert-huang transform, some feature parameters are extracted and applied to the classification of underwater target signals from actual measure. These features include (i) the center frequency of the strongest intrinsic mode function, (ii) the energy difference between the high and low frequency, (iii) the instantaneous energy variation range. Simulation and experimental results show that there exist some differences between the different types of target signals. Those may offer a good solution for automatic recognition of underwater target signals.