A statistical method for underwater target detection

This paper presents a statistical method to detect target underwater based on the statistical properties of the number of zero crossings and the amplitude of noise over a unit time interval. A new decision statistic is defined as the ratio of the mean value of the number of zero crossings to the energy of each segment when the lines appear near the lower frequency limit. Approximate expressions of the
P

_d
and
P

_fa
are derived. Experimental results taken from the sea show that the proposed method can achieve 4~7 dB suppression of the noise in comparison with the energy detection method, and can achieve 5~8 dB enhancement of the target signal in comparison with zero-crossing detection method.