Acoustic Identification Research Articles

Nonlinear factor analysis and its application to acoustical source separation and identification

Acoustical signals of mechanical systems can provide original information of operating conditions, and thus benefit for machinery condition monitoring and fault diagnosis. However, acoustical signals measured by sensors are mixed signals of all the sources, and normally it is impossible to be directly used for acoustical source identification or feature extraction. Therefore, this paper presents nonlinear factor analysis (NLFA) and applies it to acoustical source separation and identification of mechanical systems. The effects by numbers of hidden neurons and mixed signals on separation performances of NLFA are comparatively studied. Furthermore, acoustical signals from a test bed with shell structures are separated and identified by NLFA and correlation analysis, and the effectiveness of NLFA on acoustical signals is validated by both numerical case studies and an experimental case study. This work can benefit for machinery noise monitoring, reduction and control, and also provide pure source information for machinery condition monitoring or fault diagnosis.

Exploratory Study for Acoustical Species Identification of Anchovies in the South Sea of South Korea

We conducted a hydroacoustic-trawl survey from April 16–19, 2013, for acoustical identification of anchovies in the South Sea of South Korea. Three types of fish schools were defined on the basis of the trawl results: L7 (schools mixed with 57 % hairtail and 37 % anchovy), L8 (schools with 95 % anchovy), and L1 (schools with unknown species). According to the results, the bathymetric characteristics (school depth, water depth, and altitude) of L1 were similar to those of L7; however, the morphometric characteristics (length, height, area, and image compactness) of L1 were similar to those of L8. To confirm whether L1 was more similar to L8 or L7, we used the ΔMVBS method [where ΔMVBS is difference of mean volume backscattering strength (MVBS) at 38 and 120 kHz] with three cell sizes, 0.1 nmi × 2 m, 0.25 nmi × 4 m, and 0.5 nmi × 5 m; the ΔMVBS pattern of L1 seemed similar to that of L8, but not that of L7. By using one-way analysis of variance, we noted significant differences of frequency characteristics between L7 and L8 and L7 and L1, supporting the results of the ΔMVBS method. Moreover, only 0.1 nmi × 2 m and 0.25 nmi × 4 m were the appropriate cell sizes for anchovy species identification. This exploratory study may form the basis for other species identification.

Morphological and acoustic identification of Hipposiderids (Chiroptera: Hipposideridae) from Xuan Son national park, Northeastern Vietnam

Morphological and acoustic identification of Hipposiderids (Chiroptera: Hipposideridae) from Xuan Son national park, Northeastern Vietnam

Acoustic identification of Mexican bats based on taxonomic and ecological constraints on call design

Summary Monitoring global biodiversity is critical for understanding responses to anthropogenic change, but biodiversity monitoring is often biased away from tropical, megadiverse areas that are experiencing more rapid environmental change. Acoustic surveys are increasingly used to monitor biodiversity change, especially for bats as they are important indicator species and most use sound to detect, localise and classify objects. However, using bat acoustic surveys for monitoring poses several challenges, particularly in megadiverse regions. Many species lack reference recordings, some species have high call similarity or differ in call detectability, and quantitative classification tools, such as machine learning algorithms, have rarely been applied to data from these areas. Here, we collate a reference call library for bat species that occur in a megadiverse country, Mexico. We use 4685 search‐phase calls from 1378 individual sequences of 59 bat species to create automatic species identification tools generated by machine learning algorithms (Random Forest). We evaluate the improvement in species‐level classification rates gained by using hierarchical classifications, reflecting either taxonomic or ecological constraints (guilds) on call design, and examine how classification rate accuracy changes at different hierarchical levels (family, genus and guild). Species‐level classification of calls had a mean accuracy of 66%, and the use of hierarchies improved mean species‐level classification accuracy by up to 6% (species within families 72%, species within genera 71·2% and species within guilds 69·1%). Classification accuracy to family, genus and guild‐level was 91·7%, 77·8% and 82·5%, respectively. The bioacoustic identification tools we have developed are accurate for rapid biodiversity assessments in a megadiverse region and can also be used effectively to classify species at broader taxonomic or ecological levels. This flexibility increases their usefulness when there are incomplete species reference recordings and also offers the opportunity to characterise and track changes in bat community structure. Our results show that bat bioacoustic surveys in megadiverse countries have more potential than previously thought to monitor biodiversity changes and can be used to direct further developments of bioacoustic monitoring programs in Mexico.

UAV搭載マイクアレイを用いた高雑音環境下における音イベント検出・識別システム

UAV搭載マイクアレイを用いた高雑音環境下における音イベント検出・識別システム

周波数領域主成分分析に基づく多入出力構造物の音響振動モデル同定

周波数領域主成分分析に基づく多入出力構造物の音響振動モデル同定

A real-time acoustic signature-based fluid identification methodology for applications in the field of security and defence

Acoustic-based inspection methods, employed for on field screening of containers, hold high importance in the fields of security and defence. However, such methods are limited by expensive instrumentation, offline analysis and complexities related to transducer container physical coupling. Simple, non-destructive and extremely cost effective method based on acoustic resonance has been devised here for rapid online acquisition and analysis of acoustic signatures of fluids for their necessary identification and classification. The research, with two stages of experimentation, establishes reliability of the system in the first stage by using sweep excitation in fluids followed by acquiring acoustic signatures for such fluids in the second stage by using white noise for excitation. The setup uses two ceramic coated piezoelectric transducers interfaced with the computer through audio ports. Excellent consistency in the results, recorded with the help of a virtual spectrum analyzer, is achieved with average error amounting to 1.4%.

体積後方散乱強度の周波数差を利用した北海道噴火湾周辺におけるオキアミ類とカイアシ類の識別

北海道噴火湾周辺海域において，周波数 38, 120, 200 kHz の計量魚群探知機を用いて音響散乱層の特徴を調べ，その構成生物であるオキアミ類とカイアシ類の識別方法を検討した。生物の確認は，直径 80 cm のリングネットによる垂直曳網で行った。オキアミ類とカイアシ類は，周波数 38 kHz よりも 120 kHz と 200 kHz でより強い音響散乱を示し，さらにこれら 2 周波数の SV 差は，オキアミ類で−1.5～1.3 dB，カイアシ類で 2.2-3.7 dB となり，この SV 差を利用して両種を識別する可能性が示された。

Can echolocation calls of Cuban mormopid bats visualized through a heterodyne system?

In this study, we evaluate the potential of the heterodyne system for the acoustic identification of bat species from Mormoopidae in Cuba. The heterodyne transformation of the echolocation calls of the three mormoopid species of Pteronotus (P. macleayii, P. quadridens and P. parnellii) was initially analyzed by setting the frequency of the heterodyne detector (f tuned) 5 kHz above and below the constant frequency value of the second harmonic, respectively. We then studied the feasibility for the identification of several Pteronotus species with single ftuned values. The heterodyne transformation of calls from the mormoopid Mormoops blainvillii and four Cuban phyllostomid bats was studied by selecting frequency values contained in the calls of each of the species under study. We showed that, by selecting the appropriate f tuned, the Pteronotus species could be accurately identified based on the spectral signatures of their heterodyned calls. Frequency modulated (FM) bats presented very similar heterodyne signatures and therefore could not be identified to species level. The study points to heterodyne detectors as appropriate to conduct acoustic surveys of mormoopids and other constant frequency (CF) and quasi CF bat species.

Vocalisations of Killer Whales (Orcinus orca) in the Bremer Canyon, Western Australia.

To date, there has been no dedicated study in Australian waters on the acoustics of killer whales. Hence no information has been published on the sounds produced by killer whales from this region. Here we present the first acoustical analysis of recordings collected off the Western Australian coast. Underwater sounds produced by Australian killer whales were recorded during the months of February and March 2014 and 2015 in the Bremer Canyon in Western Australia. Vocalisations recorded included echolocation clicks, burst-pulse sounds and whistles. A total of 28 hours and 29 minutes were recorded and analysed, with 2376 killer whale calls (whistles and burst-pulse sounds) detected. Recordings of poor quality or signal-to-noise ratio were excluded from analysis, resulting in 142 whistles and burst-pulse vocalisations suitable for analysis and categorisation. These were grouped based on their spectrographic features into nine Bremer Canyon (BC) “call types”. The frequency of the fundamental contours of all call types ranged from 600 Hz to 29 kHz. Calls ranged from 0.05 to 11.3 seconds in duration. Biosonar clicks were also recorded, but not studied further. Surface behaviours noted during acoustic recordings were categorised as either travelling or social behaviour. A detailed description of the acoustic characteristics is necessary for species acoustic identification and for the development of passive acoustic tools for population monitoring, including assessments of population status, habitat usage, migration patterns, behaviour and acoustic ecology. This study provides the first quantitative assessment and report on the acoustic features of killer whales vocalisations in Australian waters, and presents an opportunity to further investigate this little-known population.

Advance in the bat acoustic identification systems based on the audible spectrum using nonlinear dynamics characterization

Frequential and time lineal parameters have shown a good performance in the recognition of bat species and nowadays there are many works which obtain those characteristics very accurately. However, it is necessary to move forward and test the capabilities of other characterizations on bioacoustics successfully used in other fields. In this work the chaos theory, which is an area of nonlinear dynamics systems, is applied to bat acoustic identification. The database used in the evaluation consists of 50 bat calls of seven different classes extracted from a previous work. The combinations of linear and nonlinear parameters have resulted in an average error of 1.8%, improving the accuracy in 0.42%. The differences to identify between the most difficult species and the easiest ones have been reduced.

The Problem of Low Agreement among Automated Identification Programs for Acoustical Surveys of Bats

Abstract. We compared 4 programs designed to identify species of bats from their echolocation calls (Bat Call ID, EchoClass, Kaleidoscope Pro, and SonoBat) using field data collected in Nebraska, USA (29,782 files). Although we did not know the true identity of these bats, we could still compare the pairwise agreement between software packages when identifying the same call sequences. If accuracy is high in these software packages, there should be high agreement in identification. Agreement in identification by species averaged approximately 40% and varied by software package, species, and data set. Our results are not consistent with the high accuracy often claimed by some software packages and may be a warning about the importance of understanding accuracy of acoustical identification in designing ecological experiments and interpreting results.

공명 산란 이론을 이용한 단일층 원통형 껍질 내부 물질의 음향 식별

ABSTRACT: Acoustic identification of inner materials in a single-layer cylindrical shell is investigated with acoustic resonance theory. The theoretical resonance peak frequencies for a cylindrical shell are little affected by the density variation, but remarkably changed by the sound speed variation of inner materials. Such acoustic dependency can be utilized to identify inner materials in a cylindrical shell. Acoustic resonance spectrogram for a single-layer cylindrical shell is theoretically plotted as functions of normalized frequency and sound speed of inner materials. The inner materials can be acoustically identified by overlapping acoustic resonance peaks from measured backscattering sound field on the spectrogram. To experimentally confirm this method, backscattering sound field of cylindrical shell filled with water, oil or ethylene glycol was measured in water tank. The inner materials could be identified by acoustic resonance peaks of the backscattering sound field monostatically measured with a transduce of 1.05 MHz center frequency.Keywords: Resonance scattering theory, Cylindrical shell, Acoustic identification, Acoustic resonance spectrogramPACS numbers: 43.30.Gv, 43.40.Ey, 43.40.Sk†Corresponding author: Suk Wang Yoon

False killer whale and short-finned pilot whale acoustic identification

False killer whales Pseudorca crassidens and short-finned pilot whales Globicephala macrorhynchus are known to interact with long-line fishing gear in Hawaiian waters, causing eco- nomic loss and leading to whale injuries and deaths. The main Hawaiian Islands' insular popula- tion of false killer whales is listed as endangered and the offshore population is considered 'strate- gic' under the Marine Mammal Protection Act due to relatively high bycatch levels. Discriminating between these species acoustically is problematic due to similarity in the spectral content of their echolocation clicks. We used passive acoustic monitoring along with data from satellite tags to dis- tinguish signals from these 2 species. Acoustic encounters recorded with autonomous instruments offshore of the islands of Hawai'i and Kaua'i were matched with concurrent and nearby location information obtained from satellite tagged individuals. Two patterns of echolocation clicks were established for the 2 species. The overall spectral click parameters were highly similar (22 kHz peak and 25 kHz center frequency), but false killer whales had shorter duration and broader bandwidth clicks than short-finned pilot whales (225 µs, 8 kHz (�3 dB bandwidth) and 545 µs, 4 kHz, respectively). Also, short-finned pilot whale clicks showed distinct spectral peaks at 12 and 18 kHz. Automated classification techniques using Gaussian mixture models had a 6.5% median error rate. Based on these findings for echolocation clicks and prior published work on whistle classification, acoustic encounters of false killer whales and short-finned pilot whales on auto - nomous instruments should be identifiable to species level, leading to better long-term monitoring with the goal of mitigating bycatch.

Volume backscattering strength of ice krill (Euphausia crystallorophias) in the Amundsen Sea coastal polynya

Volume backscattering strength (Sv in dB re 1m−1) of ice krill (Euphausia crystallorophias) was observed at two frequencies (38 and 120kHz) with a calibrated split-beam echosounder system in the Amundsen Sea coastal polynya. The horizontal and vertical scattering layers in the upper 200m of the water column were known with the existence of predominant ice krill (>95%) in this region. Acoustic identification using a two-frequency dB window between Sv at 38 and Sv at 120kHz separated echoes originating from dominant ice krill from other zooplankton species. The frequency dependence of ice krill at 38 and 120kHz was examined and the result presented that ice krill might have different acoustic characteristics from other Southern Ocean zooplankton species including Euphausia superba. This result could be applied to improve the ability of acoustic identification and precise density estimation of ice krill in the high-latitude coastal waters of Antarctica.

Vibration analysis and acoustic identification for a power-split planetary gear set

Vibration and acoustic characteristics of a new power-split hybrid transmission with a four-shaft planetary for passenger car are investigated in this research. A dynamic model with three degrees of freedom for the new compound planetary is established in pure electric driving mode. The numerical analysis for distinctive natural frequencies and corresponding vibration modes of this system are produced. Furthermore, in order to identify the noise source, acoustic and vibration experiments are carried out. Moreover, the finite element model for the shell cover is studied for verification of structure vibration resonant frequencies. It is demonstrated by comparison between numerical results and test data that the noise at frequency of 692 Hz is evidently related to the transverse vibration of motor E1 shaft, and the structure vibration of shell cover at frequency of 2930 Hz is exactly contributed to noise source at frequency near 3029 Hz.

Acoustic identification and taxonomic remarks of Hipposiderids and Rhinolophids (Chiroptera: Hipposideridae, Rhinolophidae) in Tam Dao national park, Northeastern Nietnam

Hipposiderids and Rhinolophids are referred to all taxa of the two families Hipposideridae and Rhinolophidae, respectively. Their echolocation signals are diagnostic by a dominant “constant frequency” (CF) segment. Each echolocation signal of leaf-nosed bat species consists of two segments (“constant-frequency” and “terminal frequency-modulated”) while that of horseshoe bat species comprises three segments (“initial frequency-modulated”, “constant-frequency” and “terminal frequency-modulated”). Between 2008 and 2014, a series of field surveys was conducted in Tam Dao National Park (TDNP) with particular emphases on taxonomy and echolocation of CF bat species. Results from the surveys indicated that TDNP is a home to three species of Hipposideridae (Hipposideros armiger, H. larvatus and H. pomona) and five species of Rhinolophidae (Rhinolophus affinis, R. luctus, R. macrotis, R. pearsonii and R. pusillus). During the present study, R. luctus was only detected in the field based on a poor echolocation call while all seven remaining species were captured. The CF of the second harmonic and morphological features of each species are clearly distinguished from the respective ones of the remaining species. This paper provides the external and acoustic diagnoses of each species for identification in the field to strengthen survey achievements, monitoring and conservation of CF bats in the park in coming time.

Analysis of humpback whale sounds in shallow waters of the Southeastern Arabian Sea: An indication of breeding habitat.

The primary objective of this work was to present the acoustical identification of humpback whales, detected by using an autonomous ambient noise measurement system, deployed in the shallow waters of the Southeastern Arabian Sea (SEAS) during the period January to May 2011. Seven types of sounds were detected. These were characteristically upsweeps and downsweeps along with harmonics. Sounds produced repeatedly in a specific pattern were referred to as phrases (PQRS and ABC). Repeated phrases in a particular pattern were referred to as themes, and from the spectrographic analysis, two themes (I and II) were identified. The variation in the acoustic characteristics such as fundamental frequency, range, duration of the sound unit, and the structure of the phrases and themes are discussed. Sound units were recorded from mid-January to mid-March, with a peak in February, when the mean SST is approx. 28 degree C, and no presence was recorded after mid-March. The temporal and thematic structures strongly determine the functions of the humpback whale song form. Given the use of song in the SEAS, this area is possibly used as an active breeding habitat by humpback whales during the winter season.

Localization for broadband source by a single vector sensor in shallow water

The physical meaning of array invariant is the absolute traveling time of different modal from a source to a receiver under the average sound speed. It can be gotten by the derivative of the arriving time against the corresponding cosecant function of elevation angle for each modal, which can be used for the passive source localization. A vector sensor can be seen as a small volumetric array. It has its capability of elevation angle determination to describe the multi-modal propagation and the single-modal dispersion in shallow water, which may imply a chance to get the array invariant for the source localization. Here, we take advantage of the dispersion based short time Fourier transform and warping transform techniques for the time-frequency analysis of vector sensor signals. This technique can improve acoustic normal mode identification and thereby extract the modal data more accurately for array invariant determination. The idea had been effectively validated during the experiment that took place in ...

Acoustic species identification of shrews: Twittering calls for monitoring

The acoustic signals of shrews (Soricidae) are largely understudied. As shrews are very vocal animals it may be feasible to use acoustic methods in field studies to assess ecological and behavioral data. In this study, we present the first detailed analysis of the twittering calls of six Central European shrew species (Sorex minutus, Sorex araneus, Neomys fodiens, Neomys anomalus, Crocidura russula and Crocidura leucodon). The analysis is based on over 6000 recorded calls from 121 individuals. Our results indicate that there is a large inter-individual variance and a large inter-specific overlap in the acoustic parameters of the calls. Each species uses a large spectral variety of calls without clear species specific call types. A species identification using the Support Vector Machine method on six species shows 66.2% accuracy; however, a pairwise comparison indicates accuracy between 68.5 and 97.3%. We propose to use acoustic monitoring of shrews in comparative studies to estimate the overall shrew activity. Moreover we suggest using the acoustic identification method in areas with few shrew species where the accuracy of the technique can be eligible.