Passive Acoustic Tracking of Cetaceans using Chirplet Transform

To study the behavior of marine mammals, dolphins, whales, and porpoises, etc., unobtrusively without influencing their behavior, a passive acoustic tracking system is used to accurately estimate the location and track of such animals. We introduce a new method to track underwater animals. We utilize chirplet transform to estimate differential delays-Dopplers between pairs of sensors of an array from the received acoustic signal, and use extended Kalman filter (EKF) for tracking. Since the acoustic signals from these animals are inherently non-stationary, the time-frequency processing of the received wavefront is required. The performance of our method is illustrated on real data, where a whale song is contaminated with simulated underwater noise in different signal-to-noise ratios.