Passive Sonar Target Detection Using Statistical Classifier and Adaptive Threshold

Abstract

This paper presents the results of an experimental investigation about target detecting with passive sonar in Persian Gulf. Detecting propagated sounds in the water is one of the basic challenges of the researchers in sonar field. This challenge will be complex in shallow water (like Persian Gulf) and noise less vessels. Generally, in passive sonar the targets are detected by sonar equation (with constant threshold) which increase the detection error in shallow water. Purpose of this study is proposed a new method for detecting targets in passive sonars using adaptive threshold. In this method, target signal (sound) is processed in time and frequency domain. For classifying, Bayesian classification is used and prior distribution is estimated by Maximum Likelihood algorithm. Finally, target was detected by combining the detection points in both domains using LMS adaptive filter. Results of this paper has showed that proposed method has improved true detection rate about 27% compare other the best detection method.

Classification of Surface Vehicle Propeller Cavitation Noise Using Spectrogram Processing in Combination with Convolution Neural Network

“…The basic approach is “one-class” classifications, in which all data from already known classes are considered a single class. According to [ 13 ], using the results of the analysis of acoustic signals on the time and frequency domains combined with Least Mean Square as input for the SVM model gives significantly improved results compared to comparable models. Yang’s research team [ 14 ] (2018) used unsorted SONAR data for pre-training to increase the effectiveness of supervised learning.…”