Passive sonar harmonic detection using feature extraction and clustering analysis

J. L. Terry, A. Crampton, C. J. Talbot

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

A current key problem in the development of passive sonar is the classification of data into its different noise sources. This paper focuses on solving the problem using feature extraction and clustering techniques. The methods described in this paper have been developed for data collected from a single sensor omni-directional passive sonar, with the input data being the extracted frequency tracks from a time-frequency lofagram. A single noise source will exhibit a number of frequency tracks within a lofagram, collectively these tracks form a harmonic set. The problem of harmonic detection is to associate the different components of each of the harmonic sets, or noise sources. Each of the frequency tracks within a harmonic set have a strong physical relationship defining their behaviour and properties. In solving the harmonic detection problem this physical relationship is exploited to associate frequency tracks with similar characteristics. In the solution to this problem each of the frequency tracks present have several key features measured. This allows the charactersitics of each of the frequency tracks to be described in a small number of key, directly comparable, parameters. In this paper a small selection of features that may be used in the analysis of the frequency tracks are described. These features are then measured for a number of different data sets. With these sets of parameterised features it is possible to associate harmonically related frequency tracks by implementing clustering analysis, since the strong physical relationship of the features mean that related tracks will be clustered. In applying clustering, decisions need to be made into how many clusters there are in the data. The approach used in this paper uses hierarchical clustering. Hierarchical clustering begins by placing each data point into its own cluster. Two clusters are then merged based on a clustering criterion. At each step two clusters are joined until all the data is held within a single cluster. The progression of the cluster can be shown in a tree diagram, or dendrogram, which is then used to find the optimal level based on the ratio of the distribution of data within a cluster and the seperation between different clusters.

Original languageEnglish
Title of host publicationProceedings of MTS/IEEE OCEANS, 2005
Volume2005
DOIs
Publication statusPublished - 2005
EventMarine Technology Society and Oceanic Engineering Society of the IEEE "One Ocean" Conference - Washington DC, United States
Duration: 18 Sep 200523 Sep 2005
http://washington05.oceansconference.org (Link to Conference Website)

Conference

ConferenceMarine Technology Society and Oceanic Engineering Society of the IEEE "One Ocean" Conference
CountryUnited States
CityWashington DC
Period18/09/0523/09/05
OtherOver 2,000 attendees from nearly 40 nations participated in over 250 events during a six-day program that included about 500 oral technical paper presentations, nearly 200 exhibit booths, 25 plenary presentations from senior leaders in Washington and throughout the community, 9 focus sessions that brought together those both familiar and unfamiliar with specific topics to engage dialogue, a wide range of tutorial sessions, special presentations from the Smithsonian Institution and National Geographic Society, a student poster competition with incredible entries, a broad spectrum of social events to encourage networking, and a first-ever Town Hall meeting on Friday morning that was the capstone event for the conference.
The conference also dedicated a special focus to education and outreach with a dynamic program that included a MATE/ROV demonstration, a full day of tutorial sessions for local educators, a plenary session on education, a presentation of the NOSB competition, a luncheon dedicated to education, a live career panel webcast for students, and the largest ever group of education-focused technical paper presentations.
Internet address

Fingerprint

Sonar
Feature extraction
Sensors

Cite this

Terry, J. L., Crampton, A., & Talbot, C. J. (2005). Passive sonar harmonic detection using feature extraction and clustering analysis. In Proceedings of MTS/IEEE OCEANS, 2005 (Vol. 2005). [1640192] https://doi.org/10.1109/OCEANS.2005.1640192
Terry, J. L. ; Crampton, A. ; Talbot, C. J. / Passive sonar harmonic detection using feature extraction and clustering analysis. Proceedings of MTS/IEEE OCEANS, 2005. Vol. 2005 2005.
@inproceedings{dfbce67e63924a25990d2045d8228df5,
title = "Passive sonar harmonic detection using feature extraction and clustering analysis",
abstract = "A current key problem in the development of passive sonar is the classification of data into its different noise sources. This paper focuses on solving the problem using feature extraction and clustering techniques. The methods described in this paper have been developed for data collected from a single sensor omni-directional passive sonar, with the input data being the extracted frequency tracks from a time-frequency lofagram. A single noise source will exhibit a number of frequency tracks within a lofagram, collectively these tracks form a harmonic set. The problem of harmonic detection is to associate the different components of each of the harmonic sets, or noise sources. Each of the frequency tracks within a harmonic set have a strong physical relationship defining their behaviour and properties. In solving the harmonic detection problem this physical relationship is exploited to associate frequency tracks with similar characteristics. In the solution to this problem each of the frequency tracks present have several key features measured. This allows the charactersitics of each of the frequency tracks to be described in a small number of key, directly comparable, parameters. In this paper a small selection of features that may be used in the analysis of the frequency tracks are described. These features are then measured for a number of different data sets. With these sets of parameterised features it is possible to associate harmonically related frequency tracks by implementing clustering analysis, since the strong physical relationship of the features mean that related tracks will be clustered. In applying clustering, decisions need to be made into how many clusters there are in the data. The approach used in this paper uses hierarchical clustering. Hierarchical clustering begins by placing each data point into its own cluster. Two clusters are then merged based on a clustering criterion. At each step two clusters are joined until all the data is held within a single cluster. The progression of the cluster can be shown in a tree diagram, or dendrogram, which is then used to find the optimal level based on the ratio of the distribution of data within a cluster and the seperation between different clusters.",
author = "Terry, {J. L.} and A. Crampton and Talbot, {C. J.}",
year = "2005",
doi = "10.1109/OCEANS.2005.1640192",
language = "English",
isbn = "0933957343",
volume = "2005",
booktitle = "Proceedings of MTS/IEEE OCEANS, 2005",

}

Terry, JL, Crampton, A & Talbot, CJ 2005, Passive sonar harmonic detection using feature extraction and clustering analysis. in Proceedings of MTS/IEEE OCEANS, 2005. vol. 2005, 1640192, Marine Technology Society and Oceanic Engineering Society of the IEEE "One Ocean" Conference, Washington DC, United States, 18/09/05. https://doi.org/10.1109/OCEANS.2005.1640192

Passive sonar harmonic detection using feature extraction and clustering analysis. / Terry, J. L.; Crampton, A.; Talbot, C. J.

Proceedings of MTS/IEEE OCEANS, 2005. Vol. 2005 2005. 1640192.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Passive sonar harmonic detection using feature extraction and clustering analysis

AU - Terry, J. L.

AU - Crampton, A.

AU - Talbot, C. J.

PY - 2005

Y1 - 2005

N2 - A current key problem in the development of passive sonar is the classification of data into its different noise sources. This paper focuses on solving the problem using feature extraction and clustering techniques. The methods described in this paper have been developed for data collected from a single sensor omni-directional passive sonar, with the input data being the extracted frequency tracks from a time-frequency lofagram. A single noise source will exhibit a number of frequency tracks within a lofagram, collectively these tracks form a harmonic set. The problem of harmonic detection is to associate the different components of each of the harmonic sets, or noise sources. Each of the frequency tracks within a harmonic set have a strong physical relationship defining their behaviour and properties. In solving the harmonic detection problem this physical relationship is exploited to associate frequency tracks with similar characteristics. In the solution to this problem each of the frequency tracks present have several key features measured. This allows the charactersitics of each of the frequency tracks to be described in a small number of key, directly comparable, parameters. In this paper a small selection of features that may be used in the analysis of the frequency tracks are described. These features are then measured for a number of different data sets. With these sets of parameterised features it is possible to associate harmonically related frequency tracks by implementing clustering analysis, since the strong physical relationship of the features mean that related tracks will be clustered. In applying clustering, decisions need to be made into how many clusters there are in the data. The approach used in this paper uses hierarchical clustering. Hierarchical clustering begins by placing each data point into its own cluster. Two clusters are then merged based on a clustering criterion. At each step two clusters are joined until all the data is held within a single cluster. The progression of the cluster can be shown in a tree diagram, or dendrogram, which is then used to find the optimal level based on the ratio of the distribution of data within a cluster and the seperation between different clusters.

AB - A current key problem in the development of passive sonar is the classification of data into its different noise sources. This paper focuses on solving the problem using feature extraction and clustering techniques. The methods described in this paper have been developed for data collected from a single sensor omni-directional passive sonar, with the input data being the extracted frequency tracks from a time-frequency lofagram. A single noise source will exhibit a number of frequency tracks within a lofagram, collectively these tracks form a harmonic set. The problem of harmonic detection is to associate the different components of each of the harmonic sets, or noise sources. Each of the frequency tracks within a harmonic set have a strong physical relationship defining their behaviour and properties. In solving the harmonic detection problem this physical relationship is exploited to associate frequency tracks with similar characteristics. In the solution to this problem each of the frequency tracks present have several key features measured. This allows the charactersitics of each of the frequency tracks to be described in a small number of key, directly comparable, parameters. In this paper a small selection of features that may be used in the analysis of the frequency tracks are described. These features are then measured for a number of different data sets. With these sets of parameterised features it is possible to associate harmonically related frequency tracks by implementing clustering analysis, since the strong physical relationship of the features mean that related tracks will be clustered. In applying clustering, decisions need to be made into how many clusters there are in the data. The approach used in this paper uses hierarchical clustering. Hierarchical clustering begins by placing each data point into its own cluster. Two clusters are then merged based on a clustering criterion. At each step two clusters are joined until all the data is held within a single cluster. The progression of the cluster can be shown in a tree diagram, or dendrogram, which is then used to find the optimal level based on the ratio of the distribution of data within a cluster and the seperation between different clusters.

UR - http://www.scopus.com/inward/record.url?scp=33947112866&partnerID=8YFLogxK

U2 - 10.1109/OCEANS.2005.1640192

DO - 10.1109/OCEANS.2005.1640192

M3 - Conference contribution

SN - 0933957343

SN - 9780933957343

VL - 2005

BT - Proceedings of MTS/IEEE OCEANS, 2005

ER -