Automatic Detection Method for Water Entry Sounds in Strong Ocean Noise Backgrounds

In a shallow sea environment with strong ambient noise, achieving high-accuracy detection of water entry sound is a challenge. Under such conditions, traditional detectors exhibit low detection rates and high false alarm rates, which no longer meet the requirements for early target detection in passive sonar. Based on the dispersion time-frequency characteristics of water entry sound propagation in shallow seas, this study proposes a detector based on Independent Component Analysis (ICA). The detector decomposes the measured spectrogram into a series of independent components and selects the one with the highest kurtosis value to form the detection function. This detector does not require prior information and only needs single-channel data. Simulation and experimental results show that the proposed algorithm achieves higher detection accuracy under the same false alarm rate and signal-to-noise ratio, with a detection gain exceeding 5.7 dB.