Passive acoustic detection and classification of divers in harbor environments

In this paper we focus on passive acoustic detection and classification of divers with open circuit breathing systems. We use a method that aims to find the fundamental breathing frequency of a diver. For the analysis we have used data from two different sea trials. One sea trial was conducted in a remote area away from most anthropogenic activities and one sea trial was conducted in the busy port of Gothenburg, Sweden. The first data set is used to characterize the signature of the divers with open breathing systems and tune the detection algorithm. We estimate the detection performance as function of varying signal-to-noise ratio levels. The detector is finally applied to data from the second sea trial that was conducted in the port of Gothenburg.