Bayesian detection and tracking of odontocetes in 3D from their echolocation clicks

Localizing and tracking of marine mammals can reveal key insights into behaviors underwater that otherwise would remain unexplored. A promising nonintrusive approach to obtaining location information is based on recordings bio-acoustic signals by volumetric hydrophone arrays. Time-difference-of-arrival (TDOA) measurements echolocation clicks emitted odontocetes be extracted used for detection, localization, in 3-D. We propose a data processing chain automatically detects tracks multiple 3-D from their clicks. First, TDOA are with generalized cross-correlation whitens received acoustic instrument noise statistics. Subsequently, tracked the domain using graph-based multi-target (MTT) method reject false close gaps missed detections. The resulting estimates then another MTT stage odontocete capability proposed demonstrated real provided two arrays recorded Cuvier's beaked whales (Ziphius cavirostris). Simulation results show presented outperform an existing relies hand annotation.