Dalarna University's logo and link to the university's website

du.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • chicago-author-date
  • chicago-note-bibliography
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Non-stationary Feature Extraction Techniques for Automatic Classification of Impact Acoustic Signals
Dalarna University, School of Technology and Business Studies, Computer Engineering.
2008 (English)Student thesis
Abstract [en]

In this work, a pattern recognition technique has been proposed to automate the process of investigating the condition of wooden railway sleeper. Condition monitoring of wooden railway sleeper is mainly performed by visual inspection and also some impact acoustics tests are manually done. Though the manual procedure uses non-destructive testing methods (visual and sound analysis), decision making is largely based on intuition; moreover the process is rather slow, expensive and also requires skilled and trained staff. Impact acoustic signals have been collected from wooden railway sleepers for the purpose of achieving automation. Given the non-stationary nature of such impact acoustics signals, emphasis in this work has only been laid on non-stationary feature extraction techniques such as Short-Time Fourier Transform and Discrete Wavelet Transform. With the help of these techniques the signals can be analyzed in both time and frequency domains. Different combinations of these techniques have been tested against classifiers such as Multilayer Perceptron, Support Vector Machine and Radial Basis Function. Data fusion was investigated on mainly two levels namely feature-level and classifier-level with an aim of getting more reliable and robust results. Experimental results demonstrate that a classification accuracy of around 84% could be achieved by fusing data at the classifier level.

Place, publisher, year, edition, pages
2008.
Keywords [en]
Short Time Fourier Transform, Discrete Wavelet Transform, Multilayer Perceptron, Support Vector Machine, Radial Basis Function Neural Network, Data Fusion
Identifiers
URN: urn:nbn:se:du-3469OAI: oai:dalea.du.se:3469DiVA, id: diva2:518460
Uppsok
Technology
Supervisors
Available from: 2008-11-01 Created: 2008-11-01 Last updated: 2012-04-24Bibliographically approved

Open Access in DiVA

No full text in DiVA

By organisation
Computer Engineering

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 158 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • chicago-author-date
  • chicago-note-bibliography
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf