Characterization of various fluids in cylinders from dolphin sonar data in the interval domain

An algorithm to map temporal signals into sequences of tinting patterns is introduced and applied to the analysis of dolphin sonar data collected during a target discrimination task. A parallel can be drawn between this algorithm and the biological auditory system which encodes sensory information in the timing patterns of neural spike trains. The information in this case is carried in the interval lengths between adjacent spikes hence we term this representation an interval domain representation (IDR). A classification feature space is defined in this domain and used to analyze the returning echoes in dolphin data. We show that the dynamical classifier is able to discriminate with high confidence cylinders filled with four different fluid types used in the experiment. The potential applications of this research include improved target discrimination and a more general classification of complex acoustic data, including biologic data. The research may also provide clues to understanding the complexities underlying dolphin sonar discrimination capabilities.