Bayesian Detection and Tracking of Odontocetes in 3-D from Their Echolocation Clicks Junsu Jang Florian Meyer Eric R. Snyder Sean M. Wiggins Simone

2025-05-06 0 0 2.41MB 57 页 10玖币

侵权投诉

Bayesian Detection and Tracking of Odontocetes in 3-D from Their Echolocation

Clicks

Junsu Jang, Florian Meyer, Eric R. Snyder, Sean M. Wiggins, Simone

Baumann-Pickering, and John A. Hildebrand

Scripps Institution of Oceanography, University of California San Diego, La Jolla,

CA 92093, USAa

arXiv:2210.12318v1 [eess.SP] 22 Oct 2022

Localizing and tracking of marine mammals can reveal key insights into behaviors

underwater that otherwise would remain unexplored. A promising nonintrusive ap-

proach to obtaining location information of marine mammals is based on recordings

of bio-acoustic signals by volumetric hydrophone arrays. Time-diﬀerence-of-arrival

(TDOA) measurements of echolocation clicks1emitted by odontocetes can be ex-

tracted and used for detection, localization, and tracking in 3-D. We propose a data

processing chain that automatically detects and tracks multiple odontocetes in 3-D

from their echolocation clicks. First, TDOA measurements are extracted with a gen-

eralized cross-correlation that whitens received acoustic signals based on instrument

noise statistics. Subsequently, odontocetes are tracked in the TDOA domain using

a graph-based multi-target tracking (MTT) method to reject false TDOA measure-

ments and close gaps of missed detections. The resulting TDOA estimates are then

used by another graph-based MTT stage that estimates odontocete tracks in 3-D.

The tracking capability of the proposed data processing chain is demonstrated on real

acoustic data provided by two volumetric hydrophone arrays that recorded echolo-

cation clicks from Cuvier’s beaked whales (Ziphius cavirostris). Simulation results

show that the presented 3-D MTT method can outperform an existing approach that

relies on hand annotation.

ajujang@ucsd.edu

I. INTRODUCTION

Passive acoustic monitoring (PAM) is a nonintrusive and eﬃcient approach for studying

and monitoring acoustically active animals, especially species that are challenging to observe

visually. PAM enables detection, localization, and tracking and is therefore well-suited

for studying abundance2,3, behaviors4, and response to anthropogenic activities5,6. With

continuously increasing human activities in the ocean7–9, consistent monitoring and assessing

the population, behavior, phenology, and physiology of marine organisms are necessary for

making informed conservation plans and management policies10. Of particular monitoring

interest are cetaceans since they are apex predators and environment sentinels11. Their

animal density and geographic location can help understand complex environmental changes,

e.g., caused by anthropogenic disturbances.

Cetaceans are known to produce various types of sounds for communication, naviga-

tion, and foraging1. They are divided into two suborders: Odontocetes (toothed whales)

and Mysticetes (baleen whales). Odontocetes predominantly use high-frequency whistles or

burst pulses to communicate12, while mysticetes produce low-frequency tonal calls, which

when used in a pattern is considered song13. To locate prey and relevant features of the

environment, odontocetes also emit echolocation clicks1, which are intense and directional

short pulses. Since clicks are impulsive and broadband in frequency as well as frequently

used underwater by odontocetes, they are promising signals for researchers to localize and

track the echolocating odontocetes with PAM.

FIG. 1. Example of acoustic source localization in 3-D from TDOA measurements. The cross is the

location of the acoustic source. Each pair of red and blue dots represents a hydrophone pair that

produces a TDOA measurement. Each TDOA measurement gives rise to a hyperboloid. The line

of hyperboloid intersection speciﬁes potential source locations. To obtain a unique source location,

further hydrophone pairs are needed.

Various PAM technologies suitable for studying whales have been developed. Promising

sensing approaches for PAM include towed arrays14, ﬁber optic cables15, mobile hydrophone

recorders16,17, and bottom-mounted hydrophone arrays4,18. Emerging accessible and in-

expensive PAM technologies are expected to provide acoustic datasets that are orders of

magnitude larger than datasets provided by conventional technologies19. Thus, establishing

eﬀective algorithmic solutions for data processing, data management, data analysis, and

performance evaluation is crucial.

This work focuses on the acoustic data recorded by volumetric hydrophone arrays that

can provide location information of echolocating odontocetes in a 3-D Euclidean space. Here,

human operators are typically required to manually inspect acoustic data, make decisions on

the presence of whales, and select promising measurements4,20,21. Fully automated tracking

of acoustically active whales from their recorded acoustic data involves numerous algorithmic

challenges. In particular, there are typically false positive detections due to noise from

the environment and the instrument itself. Furthermore, there are missed detections due

to signal directionality and signal masking by background noise. It is thus necessary to

solve a data association problem for the automated tracking of acoustically active whales

across multiple data snapshots. Data association is complicated in scenarios where multiple

acoustically active whales and other acoustic sources are present. In this work, we propose

a novel data processing chain that automatically detects and tracks odontocetes in 3-D

from their echolocation clicks and demonstrate tracking of Cuvier’s beaked whales (Ziphius

cavirostris) near the coast of California.

A. State-of-the-Art

An established method to detect acoustic signals produced by whales is to ﬁrst cross-

correlate the time series of acoustic measurements provided by a pair of hydrophones and

then apply a detection criterion to the peaks of the resulting cross-correlation signal. In

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

BayesianDetectionandTrackingofOdontocetesin3-DfromTheirEcholocationClicksJunsuJang,FlorianMeyer,EricR.Snyder,SeanM.Wiggins,SimoneBaumann-Pickering,andJohnA.HildebrandScrippsInstitutionofOceanography,UniversityofCaliforniaSanDiego,LaJolla,CA92093,USAa1Localizingandtrackingofmarinemammalscanrevealkeyi...

展开>> 收起<<

Bayesian Detection and Tracking of Odontocetes in 3-D from Their Echolocation Clicks Junsu Jang Florian Meyer Eric R. Snyder Sean M. Wiggins Simone.pdf

共57页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Bayesian Detection and Tracking of Odontocetes in 3-D from Their Echolocation Clicks Junsu Jang Florian Meyer Eric R. Snyder Sean M. Wiggins Simone

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: