Background noise cancellation for acoustic detection of manatee vocalizations.

摘要

The West Indian manatee (Trichechus manatus latirostris) has become endangered partly because of an increase in the number of collisions with boats. A device is desired to alert boaters of the presence of manatees, so that a collision can be avoided. The practical implementation of the technology is dependent on the hydrophone spacing and the range of detection. These parameters are primarily dependent on the manatee vocalization strength, the decay of the signal's strength with distance, and the background noise levels. An efficient method to extend the detection range by using background noise cancellation is proposed in this dissertation.; A Finite Impulse Response (FIR) structure and a constrained Infinite Impulse Response (IIR) structure Adaptive Line Enhancer (ALE), called Feedback ALE (FALE), that can detect and track narrowband signals buried in broadband noise are implemented in MATLAB(TM) to reduce the background noise. A library consisting of 100 manatee calls spanning ten different signal categories is established and used to evaluate the performance of the FIR-ALE and FALE with a bandpass filter acting as a baseline for comparison. A proper range of the delay for the FIR-ALE and FALE is given based on the manatee vocalizations in the established library. A field test is performed to evaluate the performance of these three algorithms in time domain using an acoustic superposition of manatee vocalizations and background noise.; The improvement of the detection range of the system enhanced by ALE over bandpass filter is estimated with lower and upper bounds. The improved detection range is also estimated using the lower bound of the performance of ALE. A real time system is developed using dSPACE. The performance of ALE is further evaluated in terms of correct detections, false detections, and missed calls. The improved SNR of manatee vocalizations may be used to extend the range of detection of manatee vocalizations and reduce the false alarm and missed detection rate during implementation of the system in the manatee's natural habitat.