Abstract: This paper proposes a new ship seismic wave signal real-time detection algorithm based on sliding wavelet energy spectrum. Firstly, wavelet-transformation is used to perform scale-wavelet energy spectrum analysis on the paired signals, and the characteristics of the signal on the scale-wavelet energy spectrum are obtained. Then, the energy in a frequency band is selected as the detection feature quantity. Finally, by comparing the feature quantity with the threshold value, and according to certain detection rules, the real-time detection of the ship seismic wave signal is completed. The measured data and simulation data show that the detection algorithm can complete the detection of the target signal in real time, and the detection probability reaches 96% when the signal-to-noise ratio is -15dB.