Target Detection in Passive Sonar with Stochastic Classifier and Adaptive Threshold

Sounds detection in the water is one of the main challenges of researchers in the sonar field. This challenge will become more complex when the depth of the sea is low or published sound is negligible. Usually Target detection in sonar system is performed with a fix defined threshold that it will increase the detection error. The purpose of this paper is to present a new target detection method in passive sonar using adaptive threshold. In the proposed method, adaptive threshold due to the statistical analysis and Bayesian classifier applied in both time and frequency domains. Particle Filter is used to calculate prior pdf in Bayesian classifier. Finally target detection point is calculated after fusion detection points in time and frequency domains by adaptive filter. The results indicate that the proposed method has better performance (23%) than recent methods in passive sonar target detection issue.