Automatic detection of vessel signatures in audio recordings with spectral amplitude variation signature

Abstract

Sound emissions by ships and boats can strongly impact marine life, with potential to affect communications, breeding and prey and predator relationships. Automatic detection of boat signatures in underwater audio recordings is thus an important task. Automated solutions are particularly relevant for monitoring preservation areas where the presence of watercrafts is usually regulated. The task is particularly challenging because it requires distinguishing multiple overlapping acoustic events in typically noisy audio recordings. In this paper, we introduce an algorithm for boat and ship detection which computes an acoustic signature that captures the variance in the frequency amplitudes observed over the duration of the signal. We evaluated the algorithm on a database of underwater recordings collected at two conservation areas in the State of São Paulo, Brazil, with very good results, and also compared it with an existing solution. Besides being effective, the algorithm requires limited user input and no parameter fine tuning to handle diverse situations. It thus provides a solution to automate the detection of vessels, with potential applications for monitoring marine preservation areas.

Deep-Learning-Based detection of recreational vessels in an estuarine soundscape in the May River, South Carolina, USA

“…More recently, passive sonar [ 29 ], acoustic tag [ 30 ], and boat engine parameters (such as shaft and engine rate, number of propellers and blades, and engine firing rate) [ 31 ] have also been utilized to detect marine vessels. Convolutional neural networks (CNN) [ 29 , 32 ] and hidden Markov models (HMMs) [ 33 ] have been applied, and Frequency Amplitude Variation (FAV) signature [ 34 ] has also been investigated in the detection process. Most recently, Wilson [ 19 ], in 2022, used a CNN method to count boats in images and further analyzed the relationship between the soundscape and the number of boats.…”

HydroMoth: Testing a prototype low‐cost acoustic recorder for aquatic environments

“…Part of the reason for the high price of existing underwater autonomous recording units ($3000-$10 000) is that they have advanced technical features and high levels of precision. These devices often include factory calibration, exceedingly low self-noise, extreme depth ratings, extensive memory and extended battery life (Sousa-Lima et al, 2013) undoubtedly valuable in some settings, there are many useful applications of PAM in aquatic habitats that do not require such expensive features, but can be completed with recordings that are uncalibrated, short in duration and/or from shallow habitats (Chapuis et al, 2021;Desider a et al, 2019;Peck et al, 2021;Reis et al, 2019). Expensive high-specification recorders are surplus to the requirements of aquatic PAM programmes like these, meaning that such programmes might benefit substantially from the development of recording systems that offered reduced capabilities at a lower price.…”

Boat Detection in Marina Using Time-Delay Analysis and Deep Learning

“…Several approaches were developed to detect the harmonic frequencies in the signals and to extract the acoustic signature of the ships. In many cases, these methods are based on the spectrum (Guo et al, 2020), DEMON spectrum (Chung et al, 2011) and Cepstrum (Das, 2013;Santos-Domínguez, 2016) and the automatic detection is usually performed by detecting the peaks with a threshold (Reis et al, 2019).…”